The research conducted by Mungas (1983) indicates that violent behavior is not a “homogeneous phenomenon” and that to try to conceptualize all violent behaviors as being qualitatively similar is misleading. He stressed that violent behavior can be best understood by attempting to identify specific patterns of behavior. Elliott (1986) proposed a biosocial model, which views aggression as the result of an interaction between the brain and the environment. Likewise, Warnken and Rosenbaum (1994) have argued that marital aggression cannot be understood by examining a single precipitating event or factor. They have approached the topic from the following areas: (a) sociocultural – environmental stressors, racial and status differences, and family interaction, (b) interpersonal – the pattern of marital and family interaction, (c) intrapersonal – personal history, personality characteristics, pathology and organic dysfunction. Raine et al. (1994) determined that the presence of a combination of contributing factors, such as described by Warnken and Rosenbaum, more than doubled the rate of violent offending in a sample that included data from over 4,000 subjects.
Demographic and Sociocultural Factors Associated with Aggression Factors
Domestic violence has been found to occur across racial and ethnic groups (Bureau of Justice, 1995). In a study conducted through the University of Texas with a sample of 555 Caucasian, 358 African American, and 527 Hispanic couples living in the United States, the frequency rate of domestic violence was reported to be 53% among African American couples, 38% among Hispanic couples, and 28% among Caucasian couples (Caetano et al., 2000). When norms regarding violence approval, age, and economic stressors were held constant, no significant differences in the likelihood of partner assault between Hispanic and Caucasian populations were detected (Kantor et al., 1994). Rates of domestic violence were found to be comparable between urban and rural locations within similar economic classes (Lee et al., 2000). Findings of a study conducted by Kantor et al. (1994) indicated that being born in the United States increases the risk of wife assault among Mexican American and Puerto Rican American husbands. A survey of Los Angeles households that included responses from 1,243 Mexican Americans and 1,149 Caucasians, reported that 21.2% of the respondents said they had been martially aggressive at least once (Sorenson, 1991). The findings of this study yielded results similar to the study conducted by Kantor et al. (1994). Spousal violence rates for Mexican Americans born in Mexico and Caucasians born in the United States were equivalent (20% and 21.6% respectively) while the frequency of marital aggression was highest (30.9%) for Mexican-Americans born in the United States (Sorenson, 1991). Police reports involving episodes of domestic violence, made between 1996 and 1997, show that Hispanics were more likely to be injured than non-Hispanics, with the likelihood of injury decreasing with each additional year of age (Duncan et al., 1999). Research conducted by Lee et al. (2002) found that African American women experience higher rates of intimate partner homicide than Caucasian women, yet the total frequency for nonfatal partner battering was similar.
Risk of violence was found to be highest in groups where violence is socially acceptable (Jewkes, 2002). The southern and western areas of the United States, associated with a greater degree of “honor related” norms that endorse violence, were found to have a higher frequency of argument-related homicide rates. These regional groups also had higher levels of violent TV viewership, violent magazine subscription rates, hunting licenses per capita and rates of execution. In addition, the laws and social policies were more favorable towards violence with looser gun control regulations, state and Congressional representatives who vote for more aggressive foreign policies and self-defense laws that are more lenient in allowing people to use violence to defend themselves and their property (Cohen, 1998).
Educational and Occupational Characteristics of Batterers
Numerous studies have been conducted with regard to battering that have found an inverse relationship between income and education. Men who engage in domestic battering are often unemployed, in a low status job, and tend to be less educated than the majority of the population (Hotaling & Sugarman, 1986). Lester (1999) has suggested that whether a person chooses internally directed aggression versus externally directed aggression might be related to their level of education. Less educated men may have a tendency to express outwardly, whereas men with higher levels of education may inhibit the outward expression of emotion and internalize aggression in an attempt to conform to the laws of society.
Straus, Gelles, and Steinmetz (1980) found that abusers often experience higher levels of job-related stress and tend to be more dissatisfied with their occupation overall. In regard to place of residence, a relationship was found between domestic violence and residence in an impoverished neighborhood, especially among African American couples (Cunradi, 2000).
In addition to frequently being at a lower socioeconomic status than the majority of the population, Bernard and Bernard (1984) reported that batterers were frequently found to be at lower levels of social status, education and financial security than their wives. Pillemer (1985) found that 64% of a group of male aggressors were financially dependent on their victims, and 55% need housing from their victims. Men who battered also were found to have significantly less education and were of a lower occupational status than men who didn’t batter (Rosenbaum et al., 1994).
Cumulatively, the results of these studies suggest that being: (1) a member of second generation racial and ethnic minority, (2) from the southern or western regions of the United States, (3) at a lower socioeconomic status, (4) financially dependent upon a partner, (5) less educated, and/or (6) from an impoverished neighborhood, increases the likelihood of becoming a batterer.
Interpersonal Aspects of Aggression
In the majority of theories of human development, both environmental factors and early childhood experience have been found to play an important role in the development of aggressive adult behavior. Fitch and Papantonio (1983) found that 60% to 80% of batterers had been physically abused by their parents, and 33% witnessed domestic violence between their parents growing up (Kalmuss, 1984). These studies support what has been referred to as the intergenerational theory, which proposes that adults often engage in behaviors that were modeled in the home by parental figures when they were children, and generalize them to other situations (Elliott, 1988; Kalmuss, 1984). Rosenbaum and O’Leary (1981) also found, in a comparison study between abusive and nonabusive husbands that abusive husbands were more likely to have been abused as children, and were more likely to have witnessed spousal abuse in their family of origin. Children often emulate the family member that they perceive to have the most power (Bandura, 1977). Violent behavior modeled by the father has been found to be particularly detrimental because the father often has the greatest power and status in traditional, more patriarchal families, and is often the primary male role model (Avakame, 1998).
Even though there is a great deal of evidence to support the premise of an intergenerational theory (Elliott 1988; Mungas, 1983; Warnken and Rosenbaum, 1994), this theory fails to explain all cases of domestic violence. Stacey and Shupe (1983) conducted a study that consisted of a sample of over 500 male spouse batterers. They reported that 40% of the participants did not witness physical violence between their parents, 60% had not been physically abused as a child, 60% had not been neglected by their parents, and 50% did not have an alcoholic father. These results suggest that there are other factors besides experiencing or witness domestic violence in the home that are responsible for the development of battering behavior.
Discrepancies in Values, Power, and Autonomy in Couples of Domestic Violence
Couples of marital violence often experience greater differences in their values, an imbalance of power within their relationship, and tend to have more dependency issues than other couples. Dissonance has often been noted between the beliefs of batters and their spouses. Hotaling and Sugarman (1986) observed that greater degrees of discrepancy in the involvement of religious activities and/or incompatibility in religious background between batterers and their wives, were associated with higher levels of marital violence (Gelles, 1974; Rosenbaum & O’Leary, 1981). Traditional men paired with nontraditional women have been found to increase marital dissatisfaction and has been associated with couple violence (Walker, 1984).
One study comparing maritally aggressive and nonaggressive men, discovered marital discord to be the strongest indicator of marital aggression (Rosenbaum & O’Leary, 1981). An imbalance of power between partners has been observed in relationships in which marital discord and domestic violence occurs. The cumulative research in this area is somewhat inconsistent, as some studies have observed that men tend to be more dominant in violent relationships, while other studies have found that women who are battered are the more assertive partners. It appears that an inequality of power, in either direction, between husbands and wives is associated with higher levels of aggression. The common ingredient among couples of marital violence seems to be an imbalance within the relationship in one or more of these areas.
Abused wives have been described as being overly passive and traditional (Lishman, 1978), and/or extremely domineering in their interactions with their abusive husbands (Snell et al., 1964). Hornung, McCullough and Sugimoto (1981) found that housewives were less likely to be in a physically violent marital relationship than women who were working, and that occupational overachievement in men was associated with a lower prevalence of spousal violence. However, they also reported that women, whose level of educational attainment was low, relative to their husbands, were at greater risk for spousal violence, especially in the form of life-threatening violence (Hornung, McCullough, & Sugimoto, 1981). Multiple studies have found that wife abuse was three times as likely when the husband dominated decision-making than when the wife dominated decision-making and about eight times more frequent than in marriages with equal power structures (Murphy & Meyer, 1991; Straus et al., 1980). Other studies have observed that in marriages of high marital conflict, female dominant relationships had the highest percentage of husband to wife violence (Coleman & Straus, 1986), and aggressors were found to be less appropriately assertive with their spouses (Rosenbaum & O’Leary, 1981).
The degree of perceived autonomy and desire for power within the relationship are also factors linked to marital violence. In a study conducted by Hotaling and Sugarman (1986), domestic violence offenders were found to be more dependent on their spouses than other men perceived differences in status by offenders in regard to their wives were found to be linked to higher levels of marital aggression. Offenders’ use of aggression was interpreted as being an attempt to enforce male dominance. Dutton and Strachan (1987) observed that maritally aggressive men scored higher on the need for power as measured by the Thematic Apperception Test, than did a group of non-batterers. Dobash and Dobash (1992) have suggested that the male aggressor’s use of marital violence is often a tactic used to maintain power, which grows out of a sense of inequality in the spousal relationship. Rosenbaum, Cohen and Forsstrom-Cohen (1991) have argued that disparities in education, income, and social status are factors that negatively impact a marital aggressor’s self-esteem. In summary, it appears that status inconsistency due to an imbalance of power and autonomy, and conflicting values within a relationship are all predictors of marital aggression (Hornung, McCullough, & Sugimoto, 1981).
Personality Characteristics of Batterers
Typologies of Batterers
Marital aggression occurs on a continuum of frequency and severity, ranging from domestic violence exclusively, to marital aggression with other violent/criminal activities. Results from research studies on batterers indicate that domestic aggression is rarely linked to chronic, generalized or severe mental disorders, but instead is linked to problematic psychological patterns that lead to violent reactions under stress (Faulk, 1974). These behavior patterns have been associated with specific personality types, or clinical profiles.
For example, Margolin and Gleberman (1988) found that batterers show narcissistic/antisocial, compulsive, and asocial/borderline features. Holtzworth-Munroe and Stuart (1994) have conceptualized three “typologies” that can be used to describe batterers. These typologies differ in several different ways, including genetic/prenatal factors, exposure to violence in the home, deviant peer relations and attitudes towards violence, impulsivity, poor social skills, attachment problems, and character disorders.
The first typology proposed by Holtzworth-Munroe and Stuart (1994) is: (a) Generally violent/antisocial batterers (generally violent), who engage in moderate to severe domestic violence, and are the most involved in extra-familial violence and criminal activity. Bernard and Bernard (1984) reported that persons like this have a strong tendency to externalize blame. They are likely to have substance abuse problems and to demonstrate antisocial and narcissistic personality patterns (Else et al., 1993; Flournoy & Wilson, 1991). The second type of batterer is: (b) Dysphoric/borderline (pathological), who engages in moderate to severe domestic violence, are the most depressed, psychologically distressed, and emotionally volatile. They are also likely to have substance abuse problems, and show borderline and schizotypal personality patterns. The third type is: (c) Family only batterers who engage in the least frequency and severity of domestic violence, and are most likely to have a passive, dependent personality pattern.
Aggression and Personality Disorders
Holtzworth-Munroe and Stuart (1994) have defined Antisocial Personality Disorder as the most severe batterer typology. Antisocial Personality Disorder includes several dimensional and categorical aspects of personality: (a) novelty seeking (b) harm avoidance, and (c) reward dependence (Cloninger, 1987). Studies with psychopaths and individuals with asocial tendencies indicate that these groups have decreased levels of arousal and decreased ability to filter stimuli, which results in stimulus deprivation, sensation seeking behaviors, and a greater desire to experience stimulating events (Raine, Lenez, & Scerbo, 1995). Volavka’s (1990) research has also uncovered a link between violent behavior and an abnormal processing of sensory input.
A second category of battering typology using the model presented by Holtzworth-Munroe and Stuart (1994) is the category associated with Borderline Personality Disorder (BPD). BPD has been defined in part as including impulsivity and frequent episodes of anger with out-of-control behavior (Valavka, 1990). Interestingly, McWilliams (1994) stated that compared to other personality disorders, individuals diagnosed with BPD tend to be extremely sensitive to rejection from others. Walker (1984) reported that about one half of all batterers threaten suicide during violent episodes with their wives, although it is unclear as to the frequency of actual attempts or completion rates (Warnken & Rosenbaum, 1994).
The third typology of battering proposed by Holtzworth-Munroe and Stuart (1994) is the family-only typology. Many of the batterers that have this typology frequently have characteristics associated with a dependent personality profile. A study conducted by Hamberger et al., 1996) found that 85% of the participants in their study fell into a passive dependent personality pattern, and that this type actually had the highest frequency of spousal violence. Based the premise that batterers often are dependent upon their spouses and strive to gain power, Waltz et al., (2000) conducted an empirical study that tested the validity of Holtzworth-Munroe and Stuart’s family-only typology (1994). The results of the study were that generally violent men tended to be more dismissing and avoidant, whereas pathologically dependent men, diagnosed with a personality disorder, were more preoccupied, ambivalent and prone to feelings of jealousy. Family-only type batterers were different from nonviolent men in that they demonstrated a compulsive care-seeking attachment style, or an anxious and ambivalent way of relating to their spouses (Waltz et al., 2000).
Dependency Issues and Styles of Attachment
West and Sheldon (1988) have described compulsive care-seeking attachment as being comprised of the three factors which are: (a) perceiving life in terms of difficulties that require help from others to resolve, (b) organizing relationships based on receiving nurturing and support, and (c) assuming that others will be responsible for one’s own needs and
decisions. Men who become domestically violent may have experienced the threat of loss of a significant attachment figure. Bowlby’s (1980) research indicates that an anger response frequently occurs in individuals threatened with loss as an attempt to prevent further separation. Ganley and Harris (1978) proposed that anger may be a critical factor in domestically violent men as feelings of hurt, fear and jealousy often appear to be immediately transformed into feelings of anger. The anger may then manifest in acts of domestic violence as a reaction against the perceived loss of an attachment figure that they are dependent upon to fulfill their needs.
Self-Esteem, Depression and Aggression
As suggested by Holtzworth-Munroe and Stuart (1994), many marital aggressors suffer from depression, in addition to Axis II disorders (Warnken & Rosenbaum, 1994). Maiuro (1988) reported that domestically violent men were more likely to be significantly depressed than generally assaultive and nonviolent men, as measured by the Beck Depression Inventory.
A number of symptoms associated with depression have been noted in batterers in various studies. Gelles (1972) found that male aggressors tend to isolate themselves, and have difficulty establishing and maintaining support systems. Goldstein and Rosenbaum (1985) reported that males who engage in marital aggression have been found to have low self-esteem and poor self-images. In addition, domestically violent men showed significantly higher levels of hostility and anger than nonviolent men (Maiuro, 1988). Rosenbaum and Bennett (1986) have proposed that depression may be linked to a sense of personal injury and anger that may lead to outbursts of temper and violence.
Other Intrapersonal Factors Associated with Aggression
Violent men were observed to misinterpret responses from their spouses more frequently, and to believe that aggressive responses would be more effective as a means to accomplishing their goals. They were found to generate less competent responses in situations with conflicts involving rejection, challenges and betrayal in the areas of communication and finance (Anglin & Holtzworth 1997).
Holtzworth-Monroe (1992) has argued that unrealistic expectations, irrational beliefs, misperceptions and processing distortions in taking in and interpreting external stimuli may lead to faulty conclusions. Erroneous interpretations may result in hostile attributions that negatively effect decision-making, which may contribute to the initiation of an aggressive response (Holtzworth-Monroe, 1992). This is supported by the results of a study conducted by Giancola and Zeichner (1994). They found that aggression was significantly related to performance on the Conditional Association Task, which measures the ability to learn a series of conditional associations between unrelated stimuli.
Coping and Anger Management Skill Deficits
In her research, Holtzworth-Monroe (1992) has identified deficits in sequential information processing in maritally violent men, linked to interpersonal interaction. Batterers often demonstrate a lack of ability in expressing themselves (LaViolette, Barnett, & Miller, 1984). In one study, maritally aggressive men, as defined by the Modified Conflict Tactics Scale (MTCS; Pan, Neidig, & O’Leary, 1994) and the Short Marital Adjustment Scale (SMAT; Locke & Wallace, 1959), were found to share the following characteristics: (a) fewer number of competent responses in spousal conflicts, (b) less likelihood of using coping skills involving reasoning, and (c) greater reliance on physical aggression. Anglin and Holtzworth (1997) found that physically aggressive spouses are less skilled in terms of problem-solving than nonviolent spouses, and appeared to have both marital and general skill deficits.
Impulsivity may refer to an abrupt inclination towards an unpremeditated action, or action without conscious and deliberate judgment (Valavka, 1995). Impulsivity has been defined to include the three following aspects: (a) Motor impulsiveness, which refers to physical activity without forethought, or acting without thinking (a primary symptom of attention-deficit/hyperactivity disorder, (ADHD), (b) Cognitive impulsiveness, which refers to acts of rapid and careless decision-making, and (c) Non-planning impulsiveness, which is a lack of planning for the future. People who demonstrate non-planning impulsivity pay most attention to events in the here and now, and tend to “discount” time in the future (Valavka, 1995).
Impulsive behavior is often associated with people who are considered to be “sensation-seeking” (Valavka, 1995), often seen in conjunction with substance abuse. Male batterers have been found to demonstrate poor impulse control and an inability to control anger (Rosenbaum & O’Leary, 1981). Fedora and Fedora (1983) found that both psychopathic and nonpsychopathic criminals with violent behavior demonstrated greater performance deficits on tasks involving executive functioning and impulse control as measured by the Trail Making Test Part B (Reitan & Wolfson, 1985) as compared to a control group. Lueger and Gill (1990) conducted a comparison study between adolescent males with and without conduct disorder (CD), which is considered to be a precursor to adult antisocial behavior. Both groups were matched on age, education, substance abuse, and socioeconomic status. They found that the CD group scored lower on the Wisconsin Card Sorting Test, which is a measure of executive functioning in disinhibition and perseveration.
Substance Abuse and Aggression
Another frequent characteristic of male batterers, also often found in people who are both impulsive and/or depressed, is a history of substance abuse. A high rate of alcoholism was reported in abusive men by their wives (Goldstein & Rosenbaum, 1985), and heavy drinkers tend to become more abusive towards their wives than moderate drinkers (Kanter & Straus, 1986). Alcohol abuse has been reported for 20% to 80% of males who engage in marital aggression, and 18% to 35% of domestic violence offenders have reported problems with drug abuse. The disinhibitory effects of drugs and alcohol can exacerbate an emotionally charged situation and increase the potential for violence (Kantor & Straus, 1987).
Intelligence and Aggression
Adult violent offenders were found to have lower Full Scale Intelligence Quotients (FSIQ) than nonviolent offenders (Spellacy, 1978). A different study found that violent offenders tended to have lower Verbal IQs as compared to Performance IQs (Yeudall, 1977). Mutschler (1997) has proposed that lower Verbal IQ scores indicate a lack of resources and planning skills necessary to process emotional states, which results in an increase in tendencies towards aggressive behaviors.
Biological Factors Associated with Aggression
As mentioned earlier, aggressive behavior patterns have been found in families for several generations with the likelihood of affecting 50% of the offspring (Elliott, 1988). Although intergenerational aggression in families may be higher due to the effects of learning through environmental exposure, this cannot be the only reason. If aggression were solely due to environmental factors, then 100% of the children raised in aggressive families would be aggressive (Elliott, 1988). Aggression is also considered to be linked to a variety of biological factors. Some of the biological factors that have been associated with aggression include hormones, organic syndromes, epilepsy, perinatal insults, birth traumas, anoxia, post-natal brain infections and diseases, exposure to toxins and genetic disorders, and sustaining a brain injury (Elliott, 1988; Kandel & Mednick, 1991; Lezak, 1995; Miller, 1999; Pennington, 1991).
Higher levels of testosterone have been considered to be a cause of increased male aggression (Elliott, 1988a). Although a positive relationship exists between testosterone levels and aggression in males (Dabbs et al., 1987). Correlation does not necessarily mean causation (Carlson, 2002). A person’s environment has also been found to directly affect their level of testosterone (Carlson, 2002). Thus, a person’s environment may cause an increase in both testosterone and aggression levels simultaneously in a parallel fashion, without the testosterone directly causing the increased aggression levels.
Genetics and Heredity Factors
Over a decade ago an argument was made that attributed violent and criminal behavior to a XXY chromosomal abnormality; however, findings were inconsistent and inconclusive (Elliott, 1988a). Although chromosomal abnormality is not present in the majority of batterers, a structural abnormality with the X chromosome and the presence of the fragile X chromosome have been found to be tied to mental impairment, which has been correlated with aggression (Elliott, 1988a).
Both the hormone and the chromosome theories illustrate the problems that can occur when trying to determine the nature of the existing relationship between hereditary factors and aggression. Pennington (2002) describes some hypotheses to explain this kind of comorbidity which include: “ (1) The two conditions share a common risk factor that is consequently not specific to either disorder; (2) One disorder may cause the symptoms of the other disorder; (3) There is an etiological subtype in which a shared risk factor produces both disorders, but that other cases of each disorder do not share risk factors; and (4) there is a nonrandom mating such that individuals with transmissible risk factors (either genetic or environmental for one disorder are more likely to have children with individuals with transmissible risk factors for the other disorder (p. 22).” The fact that we are unable to study these factors within the framework of an experimental design using random selection and assignment limits our ability to determine the exact nature of the relationship between biological factors and aggressive behavior. Because of this, current research in this area is frequently a description of common themes and shared characteristics, rather than definite causal relationships.