W Clingempeel & Eulalee Brand-Clingempeel. Handbook of Contemporary Families. Editor: Marilyn Coleman & Lawrence H Ganong. Sage Publication. 2004.

Disagreements among family members, including interparental conflicts, are ubiquitous features of family life and occur in well-functioning families (Cummings & Davies, 1994). However, a substantial body of empirical research has demonstrated that interparental conflicts exhibiting specific properties (e.g., severe, prolonged, focusing on child-rearing issues) and under specific circumstances (e.g., witnessed by children) may have a myriad of pathogenic effects on children, including a greater likelihood of internalizing and externalizing behaviors, insecure attachments, problematic sibling and peer relationships, and sustained academic difficulties (Cummings & Davies, 1994; Grych & Fincham, 1990; Katz & Gottman, 1996).

This chapter characterizes families whose interparental conflicts have negative effects on children/adolescents as pathogenic-conflict families, but given its limited scope, it will not focus on parent-child conflicts or child maltreatment separate from the influence of destructive interparental conflicts. The chapter has two major goals: first, to outline what we know, or the most robust findings, pertaining to the effects of interparental conflict/violence on children; and second, to elucidate what we need to know in three domains of inquiry: (a) cultural variations in interparental conflict and its sequalae; (b) comorbidity of spousal violence and physical child abuse; and (c) the impact of multilevel and reciprocal-influence processes on interparental conflict and its effects.

Pathogenic-Conflict Families and Children: What we Know

The effects of interparental conflict on children depend in part upon properties of the conflicts; children’s cognitive appraisals, emotions, and reactions to interparental disputes; and the extent to which couple conflicts “spill over” into parent-child, sibling, and peer relationships (Cummings & Davies, 1994; Grych & Fincham, 1990). Summaries of our knowledge in these areas are given below.

Properties of Conflicts

Continuum of Severity

Exposure to more frequent and intense episodes of interparental conflict results in greater emotional distress, psychopathology, and health problems in children (Cummings & Davies, 1994; Gottman & Katz, 1989). The more frequent and severe the conflicts, the greater is the risk that children will be adversely affected (Cummings & Davies, 1994; Grych & Fincham, 1990). Physical violence appears to have pernicious effects over and above high-intensity verbal conflicts (Cummings & Davies, 1994); and the use of potentially lethal weapons in interparental disputes (i.e., knives, guns) has greater negative effects than physical violence not involving lethal weapons (Jouriles et al., 1998).

Topic of Conflict

Conflicts focusing on parental differences on child rearing are more likely to have negative effects than conflicts focusing on other issues (Belsky & Hsieh, 1998). Children may be exposed more often to conflicts that involve them and may also be more likely to form maladaptive appraisals (e.g., self-blame attributions) of child-rearing disputes. Interparental hostilities over child rearing are more likely to engender both loyalty conflicts and cross-generation alliances two family processes linked to maladaptive child outcomes (Buchanan, Maccoby, & Dornbush, 1991; Mann, Borduin, Henggeler, & Blaske, 1990).

Conflict Exposure

Interparental conflicts that occur in the presence of children are likely to have more negative effects than those not observed by children (Davies & Cummings, 1994; Davis, Hops, Alpert, & Sheeber, 1998). Direct observation triggers negative affect and emotional insecurity (Davies & Cummings, 1994). The failure to shield children from unproductive parental disputes may reflect ineffective parenting and inadequate boundaries around marital or couple subsystems (Davis et al., 1998). Repeated exposure to interparental conflict sensitizes children to later conflicts, producing greater emotional reactivity, emotion-focused (e.g., attempts to escape or avoid conflict) and/or problem-focused (e.g., interventions designed to interrupt conflict) coping, and pessimistic expectations about the future status of the interparental relationship (Davies & Cummings, 1998; Davies & Forman, 2002). In the long run, this sensitization process increases children’s vulnerability to disturbances in psychological functioning.

Resolution Status

The manner in which conflicts are managed differentiates well-functioning and poorly functioning families better than the mere presence of conflicts (Cummings & Davies, 1994). Parents who resolve low-intensity conflicts successfully via negotiation and positive communication may have few negative effects, and may even have positive effects, on children (Grych, 1998). Children may acquire the expectation that conflicts with significant others can be successfully resolved. Alternatively, high-intensity conflicts that remain unresolved may inculcate children with the relationship message that conflict resolution in relationships is unlikely and thus may portend negatively for children’s future close relationships.

Children’s Appraisals and Reactions to Conflict

A Cognitive-Contextual Model: Children’s Appraisals

In a seminal article, Grych and Fincham (1990) proposed a cognitive-contextual framework assigning children’s appraisals of interparental disputes a significant role as mediators of the effects of interadult hostilities on children. According to this proposal, interparental conflict usually engenders both primary and secondary processing of conflict parameters. Initially, a child engages in primary processing, in which affective reactions varying in intensity signal perceived negativity and the level of threat to his or her personal well-being. If a threshold of negativity/threat is exceeded, the child engages in secondary processing, in which he or she evaluates three questions: Why is the conflict occurring (causal attributions)? Who is responsible for the conflict (responsibility attributions)? Does the child have adequate skills for coping with the conflict (efficacy attributions/expectations)? Properties of the specific conflicts, conflict histories, and characteristics of the child (e.g., age, temperament, coping skills) influence the specific appraisals. Certain types of attributions, including self-blame and inability to cope, are more likely to engender psychopathology than less threatening attributions (e.g., temporary causes of conflict, ability to cope effectively).

Aggressogenic Cognitions

Children exposed to severe interparental conflicts may acquire aggressogenic cognitions, or beliefs that support aggression as normative and a legitimate response to provocation. Aggressogenic cognitions have been found to mediate the relations between interparental conflict and children’s aggressive behavior at school (Marcus, Lindahl, & Malik, 2001). They also may shape children’s processing of social information, leading them to interpret benign behaviors of others as provocations warranting aggression. Parents’ use of aggressive conflict resolution strategies (as opposed to withdraw or discuss/negotiate approaches) have been shown to increase children’s proclivities to use aggression to cope with interpersonal disputes with peers (Dadds, Atkinson, Turner, Blums, & Lendich, 1999), and aggressogenic cognitions are mechanisms that may account for the linkage between parental and child tactics in confronting interpersonal conflict scenarios.

Specific Emotions Hypothesis

Elaborating upon the cognitive-contextual model (Grych & Fincham, 1990), Crockenberg and her colleagues (Crockenberg & Forgays, 1996; Crockenberg & Langrock, 2001) proposed a specific emotions model in which interparental conflict elicits specific negative emotions (anger, sadness, and fear) that vary depending upon both the goals that are threatened (e.g., to be loved and protected, to be able do things over which parents have control) and children’s estimations of the likelihood that goal attainment will be reinstated. Maternal and paternal marital aggressiveness are proposed to have different effects, with the behavior of the same-gendered parent having greater influence on children’s emotions and behaviors. Certain emotions are presumed to correlate with specific behavioral symptoms (e.g., fear with internalizing symptoms, anger with externalizing symptoms).

In a recent study, Crockenberg and Langrock (2001) found strongest support for the specific emotions model as applied to fathers’ marital aggression and the behavior of their sons. For boys, fathers’ marital hostility had a differential effect on behavior depending on the emotion elicited. Boys who reported fear engaged in internalizing behavior, and boys who reported anger engaged in externalizing behavior. Moreover, fathers’ marital aggression elicited greater negative emotions for both boys and girls. Certain features of fathers’ aggression (e.g., they may yell louder) may trigger more negative emotions, and paternal hostility also may signal a greater threat to goal attainment (e.g., angry fathers may be more likely to leave the home). For girls, anger, sadness, and fear increased with rising levels of fathers’ hostility.

Emotional Security Hypothesis

According to the emotional security hypothesis (Davies & Cummings, 1994, 1998), children are motivated to achieve emotional security that is threatened by sustained, high-intensity interparental conflicts. In an effort to cope with the insecurity triggered by these disputes, children may activate three interrelated processes: (a) emotional reactivity with prolonged emotional distress, (b) regulation of conflict exposure by involvement in and/or avoidance of conflict, and (c) pessimistic expectations about the meaning of the conflicts for the future welfare of the self and family. These strategies may imbue children with a transient perception of control that is adaptive during the conflict itself but is maladaptive in the long run. Exposure to destructive histories of interparental conflict sensitizes children to respond with elevated levels of these processes in response to subsequent conflicts, which, in turn, ultimately lead to adjustment problems and psychopathology. Moreover, sensitization has been shown to mediate associations between interparental conflict and children’s adjustment problems (Davies & Cummings, 1998; Davies, Myers, Cummings, & Heindel, 1999).

Children may vary substantially in the extent to which interparental conflict activates each of the three processes. For example, Davies and Forman (2002) found three distinct profiles: (a) secure children, who showed well-regulated concern and positive representations of interparental relationships; (b) insecure-preoccupied children, who both overtly and subjectively (i.e., in self-reports) evinced elevated levels of emotional distress, involvement in or avoidance of conflict, and pessimistic representations of interparental relationships; and (c) insecure-dismissing children, who displayed overt signs of heightened levels of the three component processes but self-reported low levels of subjective distress. Both profiles of insecure children reported higher levels of interparental conflict and more adjustment problems than did secure children. Preoccupied children exhibited the highest levels of internalizing symptoms, and dismissing children evidenced the highest levels of externalizing symptoms.

Children’s Responses to Conflict

Children may exhibit a variety of reactions to ongoing interparental conflict, including aggression, peacekeeping efforts, and withdrawal; and the types of reactions may influence the trajectory of interparental disputes (e.g., whether they escalate or de-escalate) and the psychological functioning of children. Aggression toward one or both parentsa common response among adolescents of both gendersmay increase the severity of the couple’s conflict and elevate the probability that the children will behave aggressively in other contexts (Davis et al., 1998). Aggressive responding toward a parent during severe interparental disputes may also increase the likelihood of physical child abuse (Appel & Holden, 1998).

Spillover into other Relationships

Spillover into Parenting

Negative moods and emotions generated by interparental conflict may spill over, or transfer, to parenting behaviors, resulting in parents’ exhibiting greater hostility and less warmth toward children (Almeida, Wethington, & Chandler, 1999). A substantial body of evidence has found support for spillover effects (Almeida et al., 1999; see Erel & Burman, 1995, for a review). At high levels of marital discord, positive parent-child relationships may be difficult to achieve (Erel & Burman, 1995). Marital conflict assessed prenatally has predicted severe physical punishment of children at 2 and 5 years following the child’s birth (Kanoy, Ulka-Steiner, Cox, & Burchinal, 2003). Moreover, consistent with a parenting-as-mediator hypothesis (Katz & Gottman, 1996), several studies have found that interparental conflict exerts negative effects on children via adverse effects on parent-child relationships (Gonzales, Pitts, Hill, & Roosa, 2000; Maughan & Cicchetti, 2002).

Interparental conflict also may have a differential impact upon mothers’ and fathers’ parenting. Several studies have found that the father-child relationship may be especially vulnerable to disturbances associated with marital disputes (Katz & Gottman, 1996; Kelly, 2000). At least two studies have found that fathers’ hostile or rejecting parenting mediated the link between interparental conflict and children’s problematic peer relationships (Katz & Gottman, 1996; Stocker & Youngblade, 1999).

Frosch and Magelsdorf (2001) hypothesized that the quality of parenting in the face of marital conflict may operate as a moderator, rather than mediator, of preschool children’s behavior problems. Consistent with their hypothesis, warm/supportive parenting buffered, or reduced, the negative effects of marital conflict on young children; whereas hostile/intrusive parenting exacerbated the negative effects of couple disputes. The authors speculated that parenting behaviors may assume a mediator role over time as the 3-year-olds in their study experience more destructive interparental conflicts.

Spillover into Co-Parenting

Interparental conflicts may also adversely affect the quality of the co-parental relationship. Co-parenting measures the extent to which parents cooperate as a team or undermine each other on child-rearing issues (McHale, 1997). Co-parenting is conceptually distinct from both the marital and the parent-child relationship (Margolin, Gordis, & John, 2001). Parents in distressed marriages, motivated by mutual desires to protect their children, may collaborate successfully on parenting issues despite hostility toward each other. Likewise, parents may have good behavior management skills and positive relationships with their children, but they may exhibit poor co-parenting by disparaging or undermining each other’s child rearing.

The early research on co-parenting focused on parents’ ability to cooperate with each other after divorce (Maccoby, Depner, & Mnookin, 1991), but more recent studies have examined co-parenting in two-parent families. For example, in newly formed stepfamilies, the presence of co-parental problems between former spouses was the most robust predictor of increases in adolescents’ externalizing behavior over time (Anderson, Hetherington, & Clingempeel, 1999). In first-marriage families, the available empirical evidence suggests that negative co-parenting may mediate the link between marital conflict and disturbances in parent-child relations and children’s adjustment (Floyd, Gilliom, & Costigan, 1998; Margolin et al., 2001).

Negative co-parenting may involve both triangulation and differential parenting mechanisms. Triangulation, or parental attempts to involve children in their disputes, may engender negative effects due to youngsters’ feeling caught between parents with whom they want to preserve positive relationships (Belsky & Hsieh, 1998; Buchanan et al., 1991). Similarly, differential parenting, or parents’ favoritism to one sibling over another, may be used to form or maintain coalitions with children that undermine or exclude the other parent. Differential parenting has been associated with negative sibling relationships and children’s adjustment problems and has been shown to contribute uniquely to children’s developmental outcomes beyond the effects of absolute levels of parental affection/control (Brody, Stoneman, & Burke, 1987; Singer & Weinstein, 2000).

Spillover into other Subsystems

Destructive interparental conflict has been associated with negative effects on sibling relationships (Brody et al., 1987; Stocker & Youngblade, 1999), the larger family system (Greene & Anderson, 1999), and peer relationships (Gottman & Katz, 1989; Stocker & Youngblade, 1999). Moreover, parental hostility toward children has been found to mediate the adverse effects of marital conflict on sibling and peer relationships, with only fathers’ hostility accounting for the negative effects of marital conflict on peer relationships (Stocker & Youngblade, 1999).

In a rare observational study of negative reciprocity in tetradic relationships (two caregivers and two children), Greene and Anderson (1999) found that, in comparison with families of nonconflictual marriages and girls, families with conflictual marriages and boys consistently exhibited greater negative reciprocity, or contingent negativity across more conversational turns without de-escalation. Girls were more likely to engage in behaviors that defused hostile interparental exchanges (e.g., peacekeeping efforts); whereas boys more often behaved aggressively, resulting in longer sequences of contingent negative reciprocity.

Pathogenic-Conflict Families and Children: What we Need to Know

Despite substantial progress, there are still significant gaps in our understanding of the diverse processes through which interparental disputes exert their effects on children. A comprehensive discussion of these gaps is beyond the scope of this chapter. Consequently, we limit our discussion in the forthcoming section to “what we need to know” in three domains of inquiry: cultural and ethnic variations in interparental conflict effects; comorbidity of interparental violence and physical child abuse; and the impact of multilevel reciprocal-influence processes on interparental conflict and its effects.

Interparental Conflict: Cultural and Ethnic/Racial Variations

Few researchers have focused on cultural/ethnic/racial variations in the effects of interparental conflicts on children. In this section, we examine how cultural variations in global belief systems, level of exposure to violence, norms regarding child socialization and emotional expressiveness, and religious practices may alter the properties of interparental conflicts, their mechanisms of influence (i.e., spillover effects), and children’s outcomes. Problems in defining culture/ethnicity/race are also discussed.

Individualism-Collectivism: Influence on Conflict Properties

Cultures vary on the global belief systems of individualism (emphasizing independence, autonomy, and uniqueness) and collectivism (emphasizing interdependence, obligations to family and the larger group, and self in relation to others), with substantial data indicating that North Americans are more individualistic than people from other parts of the world (Oyserman, Coon, & Kemmelmeier, 2002). To the extent that cultural norms emphasize harmony over independence, promote loyalty to family over personal gratification, and discourage conflict and anger, couples may engage in fewer and less severe conflicts, may be more adept at de-escalating disputes and resolving them via negotiation, may be more successful at encapsulating conflicts from children, and may cooperate better as co-parents despite animosity toward each other.

Couples in more collectivistic cultures also may experience less frequent and intense hostilities over child rearing and less parent-child conflict due to norms emphasizing family harmony and children’s obedience to parents. Korean youngsters, for example, view controlling parents as promoting the family’s welfare rather than suppressing their personal autonomy (Miller, 2002). Increased parental control is associated with greater perceived parental warmth among Korean populations but greater perceived parental hostility among European Americans (Rohner & Pettengill, 1985).

Cultural Violence: Influence on Conflict Properties

Cultures vary in the extent to which they expose individuals to violence and tolerate aggression as a solution to interpersonal problems; and these differences may be reflected in levels of interparental conflict and family aggression. Some scholars argue that the United States is a culture of violence (Straus, Gelles, & Steinmetz, 1980), and empirical evidence regarding the contributions of observing frequent media violence to subsequent aggression is compelling (Huesmann, 1999). Parents in different cultures may react differently to children’s aggressive acts, with responses ranging from explanations why aggression is inappropriate, to physical discipline, to ignoring or even reinforcing hostile behaviors (Segall, Dasen, Berry, & Poortinga, 1990). Different conflict resolution styles among couples may emerge in response to cultural forces. The level of violence exposure in the larger sociocultural context may influence the extent to which marital partners use attack, withdraw, or compromise/discuss strategies in response to inevitable disagreements, and their choice of strategies may in turn affect children’s conflict resolution strategies.

Cultural Variations in Conflict Reactions

Culture plays an important role in determining both the emotional significance of events and socially prescribed ways to communicate and act on such events (Cole & Tamang, 1998). However, little is known about how children from different cultural and racial/ethnic groups vary in their reactions to interparental conflict. Comparing two cultures of Nepali children, Cole and Tamang (1998) found that Tamang children who were ostensibly influenced by Buddhist teachings encouraging the avoidance of anger and other strong emotions were less likely to report feeling negative emotions in response to parental conflict/violence than Chhetri-Brahmin children who were not exposed to these teachings.

Cultural norms of some Asian, Indian, and West African countries discourage expression of negative emotions though tolerating disclosures of physical distress (Miller, 2002). The expression of anger may be more tolerated in cultures that emphasize the self as unique and independent (and entitled to certain rights) than in cultures that emphasize an interdependent self expected to sacrifice personal gain for the collective good (Cole & Tamang, 1998). The tendency toward passivity and behavioral inhibition (i.e., shyness) found in some Asian cultures may reduce the likelihood of overt responses to parental conflicts (Rubin, 1998). Children’s aggressive responses may be less likely in collectivistic cultures that emphasize obedience to authority and compliance with family rules. Among the Japanese, disturbances in self-esteem associated with interparental conflict may involve a diminution in culturally healthy episodes of self-criticism rather than a disruption in the optimism/self-confidence underlying the self-esteem construct in European American cultures (Kitayama, 2002). Disruptions in the attachment process stemming from interparental conflict may have different manifestations among Korean mothers and their young children than are typically found in European American parent-child relationships (Rothbaum, Weisz, Pott, Miyake, & Morelli, 2000).

Children’s emotional security in more collectivistic cultures may be distributed among more adults, and thus conflicts between biological parents may be less threatening to their personal well-being. Consequently, in comparison with European American families (which grant almost exclusive responsibility for child rearing to biological parents), children living in cultures that promote involvement of multiple caregivers in child rearing may have higher thresholds for triggering security-preserving processes in the face of severe interparental conflicts. Moreover, the relationships among the three component processes outlined in the emotional security model (Davies & Cummings, 1994) may vary across cultures.

Cultural variations in manifestations of distress at both the individual and the relationship levels suggest that widely used measures of interparental conflict, parent-child relationships, and children’s mental health based largely on European American samples may need to be recalibrated for different cultures and ethnic/racial groups in both the United States and other countries. The burgeoning diversity paradigm may require not only new ways of thinking about old constructs but completely new constructs capturing variations in family relationships currently not tapped by well-established measures (Fiske, 2002).

Cultural Variations in Spillover Effects

The extent to which spillover effects vary across cultures and ethnic/racial groups is an understudied area of inquiry and a matter of debate among family scholars. More collectivistic cultures emphasizing interdependence over independence may exhibit more spillover from interparental conflict to disruptions in parenting and sibling relationships due to the greater permeability of boundaries and interdependence of intrafamilial subsystems in these cultures (Rothbaum, Morelli, Pott, & Liu-Constant, 2000).

Moreover, for collectivistic cultures that discourage discord, conflict may be interpreted more negatively by family members and thus have greater adverse effects on multiple family subsystems (Marin & Marin, 1991). Alternatively, families from more collectivistic cultures, including Hispanic and African American families, may be buffered at least to some extent against spillover effects because of greater support from extended kin and multiple caregiver support systems (Ruiz, Roosa, & Gonzales, 2002). Thus, when spillover effects do disrupt parenting practices, the effects on children may be less negative than in European American families because more adults are involved in the child-rearing process (Ruiz et al., 2002). Moreover, families with more collectivistic orientations may be less likely to exhibit spillover because of incentives to maintain strong parent-child bonds in the face of adversity. However, at least one well-designed study (Gonzales et al., 2000) found that African American and Latino families were not immune to spillover mechanisms, as interparental conflict had adverse effects on parenting behaviors that in turn were related to children’s internalizing/externalizing symptoms.

Understanding Cultural Variation: The Definition Problem

Research on cultural variations in the effects of interparental conflict and related mechanisms of influence has been hampered by the absence of clear definitions of culture/ethnicity (Jensen & Hoagwood, 1997). Cultural, racial, and ethnic groups are not static, independent entities that cause behavioral variation in individuals and family members (Kitayama, 2002); rather, they involve dynamic processes, including socio-cultural practices, meanings, social institutions, and daily activities, that in turn may lead to variations in individual functioning. Studies comparing cultures presumed to differ on individualism-collectivism measures often have not measured directly the extent to which parental behaviors adhere to the socialization practices consistent with these global cultural belief systems. For example, studies of Asian populations often fail to measure parental attempts to promote interdependence, group harmony, behavioral inhibition, and respect for authority.

The extant research has relied too often upon transethnic, binary labels (e.g., African American, Hispanic, Asian) that obfuscate the moderating effects of national origin, level of acculturation, ethnic identification, personal and family immigration history, and religious beliefs and practices (Jensen & Hoagwood, 1997). For example, the frequency and severity of interparental conflicts may vary within specific minority families depending upon the level of acculturation or the extent to which behaviors of immigrant groups have changed toward the host culture as a result of exposure to its values, norms, and social institutions (Santisteban, Muir-Malcolm, Mitrani, & Szapocznik, 2002). Intergenerational differences on the degree of acculturation may result in interparental and parent-child conflicts. Adolescents’ rejection of the ethnic values of the culture of origin that remain important to their parents is an example of this problem (Santisteban et al., 2002). For collectivistic cultures, high acculturation within families may relate to more frequent and severe conflicts, whereas low acculturation may forecast less frequent and intense conflicts. Failure to assess the potential moderating effects of level of acculturation may obscure important sources of variation within cultural groups (Conteras, Lopez, Rivera-Mosquera, Raymond-Smith, & Rothstein, 1999).

Religion as a Cultural Variable

Future studies of cultural influences on the effects of interparental conflict on children will need to go beyond global belief systems and examine the impact of specific social institutions, including religious rituals and beliefs. Conducting a meta-analysis of 94 studies of religion and family relations, Mahoney, Pargament, Tarakeshwar, and Swank (2001) reported that greater parental religiousness was linked to less marital conflict/violence, more effectiveness at resolving marital disputes, better co-parental relationships, more consistent parenting practices, greater use of corporal punishment, and less psychopathology. Moreover, in a well-designed longitudinal study of two-parent African American families, Brody, Stoneman, and Flor (1996) demonstrated that parental religiosity had positive effects on children’s psychological adjustment via the mediating influence of less conflictual and higher-quality marital, parent-child, and family relations.

However, most studies in the meta-analysis used community samples rather than maritally aggressive couples, relied solely on self-report measures tapping global constructs, and obtained data from single informants rather than multiple sources. Thus, generalizability of the findings is limited. Furthermore, the processes by which specific religious practices and beliefs influence interparental conflict and its effect on children remain an underresearched and fruitful area for scientific inquiry in the 21st century.

Comorbidity of Spousal Violence and Physical Child Abuse

Until the 1990s, research on marital violence was segregated from research on physical child abuse, and little attention was paid to co-occurrences of violence in couple and parent-child subsystems. Although extensive research has documented that marital and parental physical aggression are independently linked to greater child psychopathology (see Slep & O’Leary, 2001), our knowledge of comorbid family violence is limited in three areas: (a) the extent of co-occurrence; (b) the nature of comorbid effects (e.g., additive, interactive, mediational, or synergistic); and (c) the individual and family-level processes leading to comorbid violence within families.

Extent of Co-Occurrence

Children who live in maritally violent homes may be at greater risk for physical child abuse. The base rate of co-occurrence among families referred to shelters for battered women or child protective services agencies for physical child abuse has been estimated at 40% (Appel & Holden, 1998). Two thirds of a sample of 232 clinic-referred adolescents who were exposed to marital aggression in the past year also experienced parental aggression (Mahoney, Donnelly, Boxer, & Lewis, 2003). The base rate of co-occurrence found in representative community samples has been estimated at 6% (Appel & Holden, 1998).

On the basis of a review of 31 studies that examined the co-occurrences of spousal violence and child abuse, Appel and Holden (1998) concluded that there is an inadequate database to evaluate the extent of co-occurrence in the United States. Significant impediments include reliance on a single informant, widely disparate definitions of marital violence and child abuse, and the absence of data from couples of different cultural and ethnic backgrounds. Most information regarding co-occurrence comes from battered women’s shelters. Few studies have collected data regarding child abuse incidents and related this information to marital violence. Longitudinal studies using more representative samples would be especially informative.

Nature of Comorbid Effects

Few studies have examined the interactive and unique effects of co-occurring marital violence and physical abuse of children. With regard to possible multiplicative effects, the results are mixed, with some studies finding that children exposed to both types of family violence experience greater psychological problems but others failing to replicate this finding (see Slep & O’Leary, 2001). With regard to unique effects (i.e., the effects of violence in one subsystem after controlling for the impact of violence in the other), there is empirical evidence that the negative effects of marital violence on children are accounted for by parental aggression against children (Mahoney et al., 2003; Maughan & Cicchetti, 2002). Less is known about the circumstances in which the links between parental aggression toward children and psychopathology are attributable to marital violence. Interestingly, Mahoney et al. (2003) obtained different results depending upon the informant. On the basis of mothers’ reports, marital aggression was not related to adolescents’ adjustment beyond disruptions in parenting. However, from adolescents’ perspectives, marital aggression was related to emotional and behavioral disturbances beyond the effects of hostile parenting behaviors.

Pathways of Influence

Appel and Holden (1998) proposed five contrasting models of the directionality of abusive relationships in families with co-occurring spousal and physical child abuse: (a) the single-perpetrator model, in which one parent operates as the sole source of violence and abuses both the spouse and child, who are passive recipients; (b) the sequential-perpetrator model, in which the victim of marital abuse is the perpetrator of child maltreatment; (c) the dual-perpetrator model, in which only one spouse is abusive toward the other but both parents abuse the child; (d) the marital violence model, in which both husband and wife are perpetrators of marital abuse and one or both parents abuse the child; and (e) the family dysfunction model, in which reciprocal aversive interactions between parents and children leading to violence are superimposed upon the bidirectional cycle of violence between husband and wife.

Various theories, including social learning, social-cognitive, developmental-ecological, family systems, and behavior genetics, may explain these contrasting pathways of influence (Appel & Holden, 1998), but to date, few studies have explicitly contrasted the models and associated theoretical explanations. In a noteworthy exception, Mahoney et al. (2003) examined the interplay of severe marital and parental aggression and adolescents’ psychopathology among a sample of clinic-referred youth. They tested two hypotheses embedded within Appel and Holden’s (1998) dual-perpetrator and marital violence models. According to the parent aggressor hypothesis, parents who direct aggression toward their partner are more likely to aggress against their offspring. According to the parent victim hypothesis, parents who are victims of marital violence are more likely to physically abuse their adolescents. Consistent with the parent victim pathway, both mothers and fathers who were targets of physical violence from partners were more likely to direct severe physical aggression toward their adolescents, even after the researchers controlled for level of marital aggression. Support for the parent aggressor pathway was obtained for fathers but not mothers. Fathers who hit their partners were more likely to direct physical aggression toward their adolescents, even after the researchers controlled for whether the men had been victims of marital aggression by wives.

Although data on base rates of the path-ways-of-influence models are unavailable, there is some evidence for reciprocal-influence processes consistent with both the marital violence and family dysfunction models. Bidirectional aggression is the most common pattern in marital violence (see Capaldi & Owen, 2001); aggression toward a parent during interparental disputes may be the most common response, at least among adolescents (Davis et al., 1998), and may increase the probability of physical child abuse (Appel & Holden, 1998). Although Mahoney et al. (2003) did not examine the family dysfunction model, their data showed that about half of families corresponded to the marital violence model, in which both parents use physical aggression in the marriage and at least one parent directs severe physical aggression toward the adolescent. A very small number of their families fit with either of the two unidirectional models, sole-perpetrator or sequential-perpetrator. However, multiple models may apply depending upon the family, and families may progress through different models over time (Appel & Holden, 1998).

Understanding Comorbid Violence: The Definition Problem

A fundamental task for researchers attempting to achieve greater understanding of comorbid family violence will be to achieve consensus on definitions, methods, and measures for assessing marital violence and physical child abuse. There is empirical support for distinguishing between the degree of severity in defining and measuring marital violence (Lawrence & Bradbury, 2001) and child abuse (Emery & Laumann-Billings, 1998), with more severe acts in both cases linked to more deleterious outcomes. Couples who engage in severe aggressive acts at the early stages of marriage are more likely to exhibit subsequent marital discord than those who engage in moderately aggressive acts (Lawrence & Bradbury, 2001). In addition, the physical and psychological impact and degree of injury resulting from aggressive behaviors as well as the type of aggressive acts should be taken into account in diagnosing marital abuse. Similar aggressive acts may have dramatically different consequences. Questionnaire measures used to classify physically abusive marital behavior typically have relied solely upon the occurrence of aggression as the diagnostic criterion for abuse, with little attention to impact and injury (Heyman, Feldbau-Kohn, Ehrensaft, Langhinrichsen-Rohling, & O’Leary, 2001).

There is no scientifically accepted consensus on what constitutes child maltreatment, including physical child abuse (Appel & Holden, 1998; Emery & Laumann-Billings, 1998). In Appel and Holden’s (1998) review, 15 different measures and/or definitions of physical abuse were used across 31 studies. Physical aggression toward children varies along multidimensional continua, including the type of behavior, intent of the perpetrator, context in which the behavior occurs, level of coercion, and community norms. The question of what is the cutoff point demarcating abuse and nonabuse and why this point is chosen rather than another is a value-laden and thorny issue. For many parents in the United States, spanking, including the use of switches and belts, is normative (Kanoy et al., 2003; Mahoney et al., 2001) and not physical child abuse, as it would be defined in some widely used measures. Distinguishing between violent and nonviolent forms of child maltreatment may have both pragmatic and scientific advantages (Emery & Laumann-Billings, 1998), but the intrusion of values and cutoff points will remain problematic.

Multilevel and Reciprocal-Influence Processes

Multilevel Conceptualizations

Children exposed to destructive interparental conflicts may exhibit substantial variability in outcomes due largely to the highly idiosyncratic interplay of protective and risk factors operating at different levels of analysis within the families’ social ecology (e.g., genetic, biological, intrapsychic, family relationships, peer group and school experiences, neighborhoods, communities, ethnicity, race, social institutions, culture). However, few studies of pathogenic-conflict families have used research designs and strategies that simultaneously assess constructs at multiple levels of analysis using multiple methods and longitudinal designs (Cicchetti & Dawson, 2002).

Multilevel research strategies may elucidate moderator effects of risk and protective factors within the broader social ecology that may attenuate or potentiate the adverse effects of interparental conflict on children. A pileup of stressors in the neighborhood and community (e.g., poverty, violence, discrimination) may increase the frequency and severity of interparental conflicts and, when conflicts occur, may also exacerbate their negative effects on children. Murray, Brown, Brody, Cutrona, and Simons (2001) found that African American mothers who perceived high levels of racism in comparison with those who perceived low levels responded to a pileup of other stressors with greater psychological distress, which in turn was linked to more pervasive disturbances in relationships with intimate partners and children. Adding risk factors to destructive interparental conflict histories may have additive, multiplicative, or exponential effects on children’s negative outcomes.

Little is known about possible protective factors that may buffer the negative effects of interparental conflicts. Although several studies have found that positive parent-child relationships may reduce the adverse effects of interparental conflict (Frosch & Magelsdorf, 2001; Katz & Gottman, 1997), researchers have given relatively little attention to protective factors in extrafamilial social environments (e.g., peer group and academic settings) or at other levels of analysis (e.g., neighborhood, community). In an exception, high levels of positive peer relationships were found to reduce the negative effects of violent marital conflict and harsh discipline on children’s externalizing behavior (Criss, Pettit, Bates, Dodge, & Lapp, 2002). Academic achievement and concomitant positive relationships with teachers and school personnel consistently emerge as factors promoting resilient outcomes in high-risk children (Masten & Coatsworth, 1998), but few studies have examined children’s cognitive competence and characteristics of the school environment as possible protective factors that may reduce specifically the adverse effects of interparental conflicts and violence on children.

Interparental Conflict: Reciprocal-Influence Processes

Most studies of the effects of interparental conflict on children have focused on unidirectional models. Extensive research has documented that interparental conflict has direct negative effects on children’s adjustment as well as indirect adverse effects via the mediating influences of disruptions in parent-child and co-parental relationships. However, plausible bidirectional or reciprocal-influence processes rarely have been studied. Children’s psychiatric symptoms (e.g., externalizing behavior) may engender interparental conflict and physical child abuse (Appel & Holden, 1998). Problematic parent-child relationships including harsh punishment and abuse may trigger interparental disputes escalating to marital violence (Belsky & Hsieh, 1998). The bidirectional models proposed by Appel and Holden (1998) to account for co-occurrences of interparental violence and physical child abuse (e.g., negative reciprocity between parents and/or parents and children may escalate to violence) have received relatively little research attention. Children’s conflict with siblings and peers may aggravate parents, leading to more hostility in parent-child and marital relationships (Stocker & Youngblade, 1999).

The relative strength of effects of bidirectional pathways may vary depending upon family structure and stage of the family life cycle. For example, in the Hetherington and Clingempeel (1992) longitudinal study of remarriage, interdependencies of marital conflict, parent-child relationship problems, and adolescents’ externalizing behavior differed for newly formed stepfamilies (custodial mothers remarried 4–26 months) and well-established first-marriage families (biological parents married 9–15 years). In first-marriage families, the data strongly supported the pathway from marital negativity to problematic parenting to adolescent externalizing behaviors. In contrast, in stepfamilies, the results indicated a pathway from adolescent externalizing behaviors to negative parenting/stepparenting to marital conflict. Thus, in first-marriage families, marital conflict was the driving force, whereas in stepfamilies, adolescent-driven effects were prominent. Reciprocal-influence processes also may vary within and across cultural/ethnic/racial groups, but so far this topic has received little research attention.

Conclusion

Family researchers have amassed an impressive body of knowledge regarding the effects of interparental conflict on children. Many features of Grych and Fincham’s (1990) cognitive-contextual model have received empirical support, and refinements of their seminal framework, including the specific emotions model (see Crockenberg & Langrock, 2001), the emotional security hypothesis (Davies & Cummings, 1994; Davies, Harold, Goeke-Morey, & Cummings, 2002) and documentation of spillover and parenting-as-mediator effects (Erel & Burman, 1995; Maughan & Cicchetti, 2002), have enhanced our understanding of pathogenic-conflict families. However, many questions in the three domains of “what we need to know” discussed in this chapter remain unanswered. Twenty-first-century family scholars who address these questions will confront substantial theoretical, methodological, and data analytic challenges; but a greater understanding and appreciation of human diversity in all its forms will be the valuable fruits of their labor.