There is a considerable body of evidence to suggest that the cycle of violence hypothesis logically explains how aggressive behaviours are transmitted and maintained within certain families. In the past, researchers hypothesized the existence of a causal link between physical abuse experienced in childhood and later adult violence. However, research in this area has failed to illustrate a definitive cause and effect rule explaining the intergenerational transmission of violence. It is becoming more apparent, through evidence from current research, that the learning of aggressive behaviour is a complex process in which the interactions between individual, environmental and structural variables produce different outcomes in different individuals. The result of this complexity is a profound difficulty determining which variables interact to increase the potential for violent behaviour in adulthood. Although it seems reasonable to assume that children learn to become violent by being subjected to violence through physical abuse or by watching others commit abusive acts; presently, the evidence to support the cycle of violence hypothesis is weak and often overestimated.
The cycle of violence hypothesis is problematic because it does not explain why, in some instances, parental neglect or rejection causes more dysfunction in behaviour than physical abuse (Widom, 1989a; McCord, 1991). In addition, there is no overwhelming evidence to suggest that parents who are abusive were themselves abused as children (Widom, 1989b).
Another serious challenge for the cycle of violence hypothesis is that most children who are abused or neglected do not go on to become offenders. In a study by Widom (1989a), it was found that the occurrence of abuse or neglect did increase one's risk for delinquency and criminal behaviour; but 74% of abused or neglected subjects had no juvenile offenses and 89% had no arrests for violent criminal acts.
The cycle of violence hypothesis also disregards the potential of violence to produce other dysfunctions in behaviour, such as, withdrawal or self-destructiveness. Widom (1989a) suggests that there may be a "more subtle emotional damage" (Widom, 1989a, 164) that occurs with childhood abuse or neglect and that the resulting behavioural manifestations are dependent on the presence or absence of other factors. For example, there are more women than men who suffer from depression and undergo psychiatric hospitalization, and this may reflect gender differences in socialization that function to discourage women from outwardly displaying aggression.
These problems with the cycle of violence hypothesis have led researchers to try different approaches to identify which specific family variables are more likely to influence the transmission and maintenance of aggressive behaviours.
Haapasalo and Tremblay (1994) studied young boys with different patterns of fighting behaviour to see how these groups differed from each other in family background and later delinquency. The results showed that the stable high fighters had a more socially disadvantaged family background. Early aggressive behaviour appeared to be correlated with poor childrearing practices, measured by level of supervision, number of rules and level of punishment; but the authors note that these practices could have been a reaction to previous aggressive behaviour. Early high fighting patterns were found to be associated with a high level of delinquency, particularly if high fighters were also in the high-punishment subgroup. The authors concluded that the family process variables of perception of parental supervision and punishment were more important to the development of later aggressive behaviour than family context variables.
The Bloomington Longitudinal Study (Bates, Bayles, Bennett, Ridge & Brown, 1991) attempts to discover the early antecedents of childhood aggression. The importance of this study is not in identifying biological determinants of behaviour problems, but in emphasizing how parent-child interactions are reciprocal events. The authors found some links between the mother-perceptions of infant temperament and later behavioural adjustment. The results showed that more positive maternal involvement and less negative control was linked with the presence of fewer behaviour problems. The authors suggest that this effect is due to the infants being taught a positive sense of self and learning more socially valued behaviours which would prevent the maintenance of inappropriate behaviours. Conversely, when an infant's difficult temperament elicits a more negative response from the mother, this type of conflict interaction is more likely to be maintained throughout childhood.
Another study (Snyder & Patterson, 1995) also examined parent-child interactions. The authors applied coercion theory to show how the process of reinforcement could explain individual differences in aggression. From this perspective, the maintenance of aggressive behaviour is achieved by its "functional value" (Snyder & Patterson, 1995, 372) in preventing others from controlling one's own behaviour. It was found that aggressive mother-child dyads were more likely "to reinforce each others' aggressive means of dealing with conflict" (Snyder & Patterson, 1995, 388). These interactions should provide children with the opportunity to develop skills for dealing with disagreement with others. Children in nonaggressive dyads learned strategies that are more likely to result in conflict resolution; whereas, children in aggressive dyads leran to escalate conflict until one individual is outcoerced. From these results, the authors suggest that the treatment of childhood aggression should involve both child and parent. Each must be taught to respond positively to the use of constructive tactics. Also it is important that aggressive tactics are ignored or receive a more constructive response to avoid negative reinforcement of aggression. Once aggressive tactics lose their value they will begin to be replaced with more constructive responses that have increased value for the individual.
In a longitudinal study by Farrington (1991) measures of aggression were reported at ages 8, 10, 12, 14, 16, 18 and 32, for a group of male subjects. Violent and nonviolent offenders were compared to determine if any significant differences were apparent in the developmental pathways of these particular types of antisocial behaviour patterns. Violent offenders differed significantly from nonoffenders on many factors in childhood, adolescence and adulthood, but did not differ significantly from and were essentially the same as nonviolent offenders (Farrington, 1991). The author concluded that subjects showed a continuity in aggressive behaviour throughout their lifespan, but that this trend was only one aspect of a more "general continuity in antisocial and deviant behaviour" (Farrington, 1991, 25). From this data, Farrington suggests that the cause of violence should be traceable to the same cause of "persistent and extreme antisocial, delinquent and criminal behaviour" (Farrington, 1991, 25).
Eron, Huesmann & Zelli (1991) reviewed the findings from two longitudinal studies investigating the effects of parental variables on the learning of aggression. Parental rejection, punishment for aggression and lack of identification of the child with the parent were associated with aggressive behaviours in childhood and adulthood. However, these parental variables were also found to be probable responses to the child's aggressive behaviour and not necessarily the cause of that behaviour. It was found that when the aggression level in boys at age 8 was controlled for, the probable causal effect of parental variables disappeared. Yet, for girls, there was a slight effect, whereby those who were more severely punished at 8 years were more likely to use harsh punishment with their own children and to be more prone to aggressive responses with others generally. Finally, it was concluded that the most reliable predictor of aggression in adult life was the presence of early aggression. The authors suggest that although the parental variables appear to have no effect on the development of aggression, or if they do that effect must occur prior to 6 years of age, that parent responses to aggression may still play a role in sustaining an aggressive behaviour pattern. For example, parents have the power to intervene directly to reinforce aggression or to teach more constructive behaviours.
What can be determined about the development of aggression from the literature reviewed here? Aggression can result from parental rejection, neglect or abuse, yet it does not develop in all individuals, nor do these parental behaviours affect all individuals similarly. When aggression does appear it tends to evolve very early in the developmental process and has been shown to remain fairly stable through the lifespan of an individual. Additionally, if aggresive behaviour is present, it is usually only one aspect of a much larger antisocial pattern of behaviour. If childhood aggression is the most reliable predictor of later aggression, what is happening to encourage this behaviour and what makes it so difficult to extinguish? Although it seems that parental variables have very little effect on the initial development of aggression; is it possible that these variables play an important role in its maintenance? One explanation is offered by McCord (1991) who suggests that the critical element in establishing a lifelong pattern of aggression is the use of punishment as a response to agression and other negative behaviours.
The main argument against the use of punishment is that it provides a model for violence which the child generalizes to include illegitimate uses of force as acceptable behavioural responses (McCord, 1991). According to the theory of cultural spillover, the intergenerational transmission of violence occurs because physical punishment and physical abuse teach norms of violence which are observed and accepted by children (McCord, 1991). However, previously reviewed data from Widom (1989a), shows that child neglect and rejection produce as much and, in some cases, more violence in later life. In these situations, socialization is more likely to be weak or incomplete and thus the acceptance of norms is a questionable explanation (McCord, 1991). Furthermore, McCord refers to studies that demonstrate how parental warmth and affection enhance the acceptance of norms. If parents are affectionate and punitive one would expect to see an increase in the acceptance of norms of violence and if parents were punitive, but not affectionate one would expect a rejection of these norms; instead the reverse is true (McCord, 1991).
McCord suggests that the data can be explained more adequately by construct theory. She proposes that the same mechanism that relates parental physical abuse and punishment to later expressions of aggression can be extended to include nonphysical punishments and neglect.
First, it is necessary to examine how children learn. Infants and young children are not purely egocentric in their behaviours; but rather, they acquire knowledge from the environment and use it to guide their behaviour. For example, a child's perception of pain can be influenced by whether or not another person is present, simply because the observer's reaction informs the child the she or he should ffel and show pain (McCord, 1991). Secondly, children learn accepted behaviours and beliefs while acquiring language. As their knowledge increases, children construct classification systems which exert an influence, beyond the learning of grammar and vocabulary, to what is paid attention to and how one should behave (McCord, 1991).
At an early age, children learn the basic reasoning behind punishment ("if....then....") while they are learning to create sentences (McCord, 1991). Although punishment is intended to convey the message to stop a particular behaviour; other messages are indirectly transmitted (McCord, 1991). For example, if a punishment is going to be effective it must be intended to cause some kind of pain and the child is shown what the parent believes is painful. McCord states that this can create an ironic dislike for the very activities which parents wish to encourage; such as, schoolwork and doing chores (McCord, 1991). Furthermore, the child is aware of the punisher's intention to cause pain. This may decrease a child's desire to affiliate with the punisher and, thereby, reduces the punisher's influence over the child's socialization (McCord, 1991).
McCord also cites the results of studis which show that severe threats actually increase the value of 'the forbidden' and transmit the message that certain things are more valuable than others. Consequently, when a severe threat is issued children look for what is valuable about that which is forbidden and make a decision about how to behave based on their own self-interest. Thus, through the use of punishment, children learn to monitor their behaviour according to what they have to gain versus what they might lose. Conversely, in the mild threat condition, children lower the value of the 'forbidden' object based on a "concern about how the experimenter might feel" (McCord, 1991, 174).
The use of rewards to encourage particular behaviours can have similar negative effects. Rewards also transmit messages about what things should be valued and have been shown to provoke the devaluation of the behaviours or activities which were meant to be encouraged (McCord, 1991). Consequently, the use of punishment and reward both function to teach children to use estimates of their own "self-interest as the legitimate grounds for choice" (McCord, 1991, 176).
Parental neglect and rejection teach children directly about self-interest. They learn from an early age that they are responsible for meeting their own needs and they watch the adults around them acting in their own self-interest; instead of in the interest of their children. Furthermore, without affectionate parents to teach them; children have no method of learning more socially approved norms to guide their behaviour. In this way, children who are abused, neglected or punished are not adequately exposed to norms, such as, social responsibility, reciprocity and equity, which would promote prosocial behaviours. However, they do internalize a "norm of self-interest" (McCord, 1991, 170) which directs their behaviour into an antisocial pattern, in which aggression is only a device which functions to increase one's power to benefit in a variety of situations.
In sum, construct theory provides a way of reconciling some of the inconsistent data about the cycle of violence. Aggressive behaviours can be facilitated by the experience of parental abuse, neglect, rejection and punishment through the transmission of a "norm of self-interest" (McCord, 1991, 170). This process supports the previous data that aggression has early beginnings and is part of a larger antisocial pattern of behaviour. Aggression is only one aspect of this pattern and is perceived as having a functional value in that it serves one's own self-interest when necessary. The process also supports the data on the continuity of aggressive antisocial behaviour throughout the lifespan. The reinforcement contingencies present in the developmental environment would have to change and begin to reinforce other behavioural options before the aggressive behaviour could be extinguished. It is unlikely that the family environment of a physically abused, neglected or rejected child would provide enough positive reinforcement of behaviours other than aggression and self-interest. Yet, it is encouraging that a majority of children from abusive homes do not go on to commit violent acts; and therefore, it would be important, in future research, to distinguish the specific protective factors which exist for these individuals. Once these factors are recognized they can be used to help change the future for other children in violent family atmospheres. It is only through a better understanding of the processes involved in the intergenerational transmission of violence that families at risk can be effectively identified and successful intervention programmes can be developed.
References
Bates, J., Bayles, K., Bennet, D., Ridge B. & Brown, M. (1991). Origins of externalizing behavior problems at eight years of age. In Pepler D. & Rubin, K. (eds.). The Development and Treatment of Childhood Aggression. Hillsdale, NJ: Lawrence Erlbaum.
Eron, L., Huesmann, L. & Zelli, A. (1991). The role of parental variables in the learning of aggression. In Pepler, D. & Rubin, K. (eds.). The Development and Treatment of Childhood Aggression. Hillsdale, NJ: Lawrence Erlbaum.
Farrington, D. (1991). Childhood aggression and adult violence: Early precursors and later life outcomes. In Pepler, D. & Rubin, K. (eds.). The Development and Treatment of Childhood Aggression. Hillsdale, NJ: Lawrence Erlbaum.
Haapasalo, J. & Tremblay, R. (1994). Physically aggressive boys from ages 6 to 12: Family background, parenting behavior, and prediction of delinquency. Journal of Consulting and Clinical Psychology, 62:5, 1044-1052.
McCord, J. (1991). Questioning the value of punishment. Social Problems, 38:2, 167-176.
Snyder, J. & Patterson, G. (1995). Individual differences in social aggression: A test of a reinforcement model of socialization in the natural environment. Behavior Therapy, 26, 371-391.
Widom, C. (1989a). The cycle of violence. Science, 244, 160-165.
Widom, C. (1989b). Does violence beget violence? A critical examination of the literature. Psychological Bulletin, 106:1, 3-28.
return to the social psychology page
return to the main psychology page
