Social learning theory – is one of the most popular explanatory perspectives in the marital violence literature. Often conceptualized as the ‘‘cycle of violence’’ or ‘‘intergenerational transmission theory’’ when applied to the family, the theory states that people model behavior that they have been exposed to as children. Violence is learned through role models provided by the family (parents, siblings, relatives, and boyfriends/girlfriends), either directly or indirectly (i.e., witnessing violence), is reinforced in childhood, and continues in adulthood as a coping response to stress or as a method of conflict resolution (Bandura 1973).
- Research Supporting the Intergenerational ‘‘Cycle of Violence’’ Theory
- Intergenerational Transmission and Gender
- Sex-Role Theory
- Is Aggression Generalizable?
- Mediators of Childhood Exposure to Violence and Intimate Partner Violence
- Prevention Implications
During childhood and adolescence, observations of how parents and significant others behave in intimate relationships provide an initial learning of behavioral alternatives which are ‘‘appropriate’’ for these relationships. Children infer rules or principles through repeated exposure to a particular style of parenting. If the family of origin handled stresses and frustrations with anger and aggression, the child who has grown up in such an environment is at greater risk for exhibiting those same behaviors, witnessed or experienced, as an adult. Gelles (1972) states that ‘‘not only does the family expose individuals to violence and techniques of violence, the family teaches approval for the use of violence.’’ Children learn that violence is acceptable within the home and is an effective method for solving problems or changing the behavior of others.
The primary hypothesis for the intergenerational cycle of violence is that violent and abusive adults learned this behavior as a result of being the victims of or witnesses of aggressive and abusive behavior as children. If children are abused by their parents, they may internalize beliefs and patterns of behaviors that lead them to abuse their own children; if children observe parents who hit each other, they may develop a greater propensity toward abusing their own spouses. Transmission of violent behavior occurs through processes ofmodeling, failure to learn appropriate ways to manage conflict, and reinforcement for violent behavior. Normal coping mechanisms may not be learned or may become impaired, leading to violence as the ultimate resource.
1. Research Supporting the Intergenerational ‘‘Cycle of Violence’’ Theory
There are numerous studies that support the cycle of violence theory, showing that the experience of violence in childhood is associated with general patterns of violent behavior (Widom 1989), as well as later violence in one’s intimate relationships (Browne 1980; Burgess, Hartman, and McCormack 1987; Fagan, Stewart, and Hansen 1983; Gelles 1972; McCord 1988; Roy 1982; Steinmetz 1977; Straus, Gelles, and Steinmetz 1980; Walker 1984). Early support for the cycle of violence was buttressed by two reviews of the literature. A review of findings from six studies (Okun 1986) indicated that 23 to 40 percent of battered women witnessed violence between their parents, while in four studies 10 to 33 percent of battered women were also abused as children. Hotaling and Sugarman (1986) reviewed fifty-two case comparison studies of marital violence, finding that witnessing violence between parents was a consistent risk marker for spouse abuse among both males and females. Although not a consistent risk marker, the majority of studies also found an association between being a victim of childhood violence and spouse abuse.
Much of the early work on intergenerational transmission was derived from small cross-sectional studies of distinctive populations, such as clinical populations and children of battered women in shelters. Appropriate control group comparisons were often missing, making it difficult to establish cause and effect. Results from these studies have generally supported the association between witnessing or experiencing violence in childhood with later negative outcomes, such as partner violence. The linkage is somewhat less pronounced in nonreferred, community samples (Margolin 1998; Stith et al. 2000). This may be a result of the less severe nature of most violence that occurs or a result of better controls. Women in shelters report an average of sixty-five to sixty-eight assaults per year, which is about eleven times greater than the average of six assaults per year reported by abused women in the National Family Violence Survey (Straus 1990a). Other limitations of the early studies included the use of retrospective data. Retrospective assessments rely on long recall periods, with the possibility of selective recall biases and memory reconstruction problems. Discrepant findings using retrospective versus prospective designs have been documented. For instance, a study that examined whether childhood victimization increased the risk for drug abuse in young adulthood found increased risk with retrospective self-reports, but no risk when prospective data was used (Widom, Weiler, and Cottler 1999). A better approach for studying the linkage between early exposure to violence and later partner violence is to utilize longitudinal studies.
Overcoming the issues of retrospection and lack of comparison groups, White and Widom (2003) used a prospective study to trace long-term outcomes for men and women with official records of child abuse and/or neglect prior to age twelve, and a control group of nonabused matched cases. Both groups (n = 939) were interviewed twenty years later to discover that the abused and neglected children were slightly more likely than controls to ever hit their partners (53 vs. 41 percent). This difference held for both males and females.
Another twenty-year prospective study using a randomly selected sample of youth and their mothers residing in two upstate New York counties in 1975 followed 543 children to test the independent effects of parenting, exposure to domestic violence between parents, maltreatment, adolescent disruptive behavior disorders, and emerging adult substance abuse disorders on the risk of violence to and from an adult partner (Ehrensaft et al. 2003). Consistent with social learning theory, observing violence between parents, childhood powerassertive punishment by the mother, and adolescent conduct disorder (which appeared to mediate the effects of childhood physical abuse) were predictors of later perpetration of partner violence. Observing parental violence also predicted later victimization by a partner. Childhood physical abuse significantly predicted injury by a partner, as well as injury to a partner.
Support for the theory that direct or indirect (i.e., witnessing) childhood exposure to parental violence is related to engaging in later partner violence is also supported by nationally representative samples, such as the 1975 and 1985 National Family Violence Surveys (Straus 1990b; Straus et al. 1980). Relying on retrospective data, Straus found that males and females who endured more (i.e., higher frequency) ordinary physical punishment as children had higher rates of both ordinary and severe marital violence as adults. They also reported higher rates of ordinary physical punishment and child abuse toward their own children. Men and women who had witnessed parents hit each other were three times more likely to abuse their own partners compared with those who had not. Respondents with the experience of being both abused as children and witnessing parental violence— the ‘‘double whammy’’—had a one in three chance of encountering marital violence in the study year, double the overall rate for annual marital violence. Subsequent analyses confirmed that dually exposed, compared with singly exposed, women had significantly increased risk for adult perpetration of child abuse and for partner abuse perpetration and victimization. Similarly, men exposed to both forms, rather than one form, of family-of-origin violence had double the risk of partner abuse victimization. Men’s risk for perpetration of child abuse or partner abuse was elevated by exposure to any form of family-of-origin violence but was not increased by exposure to multiple forms of family-of-origin violence.
As evidence mounted in support of the cycle of violence theory, a new criticism arose that studies failed to separate witnessing violence from experiencing violence. These two types of exposure to violence may differentially affect the learning of marital violence. Kalmuss (1984) explored the relationship between childhood family aggression (by those children who directly experienced violence and those who only witnessed it in their families) and severe marital aggression in the next generation, using data from 2,143 adults in the 1975 National Family Violence Survey. In this retrospective study, she found that severe marital aggression was more likely when respondents, males and females, observed hitting between their parents than when they were hit as teens by their parents, although both forms of first-generation violence resulted in increased levels of second-generation marital aggression. Exposure to both types of childhood aggression led to a dramatic increase in the probability of marital aggression.
Recent support for the intergenerational cycle of violence theory comes from meta-analysis, which is a systematic review of the relevant literature allowing for statistical aggregation of results that can be reported as an average effect size. A meta-analysis of thirty-nine studies that examined the relationship between witnessing or experiencing family violence in childhood and receiving or perpetrating violence in an adult heterosexual cohabiting or marital relationship demonstrated that growing up in an abusive family is positively related to becoming involved in a violent marital relationship (Stith et al. 2000). The relationship is weak to moderate, with r values ranging from .08 to .35, depending upon the relationship examined.