Toy stories: aggression in children's narratives in the United States, Sweden, Germany, and Indonesia.

One hundred twenty 4-year-old children (30 from each culture, half girls) told two stories using toys with aggressive and neutral cues. Preschool teachers rated children's social competence and classroom behavior and parents completed a questionnaire about their child-rearing practices. Children's narratives were coded for length, complexity, story characteristics, thematic content, and number of aggressive words. Results showed that American children's narratives had more aggressive content, aggressive words, unfriendly characters, and mastery of situations with aggression than those of the Swedish, German, and Indonesian children. Although there were cross-cultural differences in the frequency of aggression in the narratives, there were similar intracultural patterns in children's individual characteristics that were related to aggression in the stories. The findings suggest that children's narratives reflect their knowledge and experience of aggression in their culture.

Previous research on the thematic aspects of children's fantasies expressed in either imaginative play or story narratives suggests that pretend scenarios reflect children's knowledge of real life events (Bretherton, 1984; Garvey, 1990), their concerns (Fein & Rivikin, 1986), and attempts to organize and make sense of their experiences (Bruner, 1990; Engel, 1995; Nicolopoulou, Scales, & Weintraub, 1994; Paley, 1986). In constructing narratives, children incorporate personal experiences, knowledge of people, social interactions, and current events into culturally available images and symbolic frameworks. Narratives also provide opportunities for children to express and symbolically resolve troubling emotional issues (Nicolopoulou, 1993).

Play and story narrative techniques have provided valid and reliable ways to measure children's understanding and expression of aggression, particularly among young children who may not respond to direct questioning or have difficulty articulating their thoughts and feelings. For example, in a study examining children's story narratives after the Los Angeles riots of 1992, Farver and Frosch (1996) found that children who had direct exposure to the riots told stories with more aggressive content than a control group of children who had no direct riot exposure.

More recently, results from two pilot studies conducted with American preschool children showed that their story narratives had highly aggressive content despite the type of toys used (aggressive vs. prosocial), their social class (middle- and low-income), or ethnicity (Anglo-American, Mexican American, and African American) (Kim, 1996; Locke, 1995). Furthermore, the findings indicated that children who told stories with highly aggressive content were also verbally and physically aggressive with peers during observations of their free play activities, were rated by teachers as being difficult socially, and had mothers with restrictive parenting styles. The present study builds on this prior research by examining preschoolers' story narratives in the United States, Sweden, Germany, and Indonesia to explore individual differences in children's expression of aggression and to understand how culture may influence this expression.

There is no current research on children's narratives in other societies and there are few studies that have examined aggression in young children cross-culturally. However, one would predict that there would be differences in the aggressive content of children's narratives based on the socioculturai contexts in which they are raised. Although aggression is possibly a universal human phenomenon, comparative studies show that there are vast cultural differences in the tolerance and prohibition of aggressive behavior, the frequency of aggressive acts, and incidents of violent crime (Goldstein & Segall, 1983). According to a model proposed by Segall (1983), ecocultural variables such as prevailing beliefs about violence and aggression, norms governing conflict resolution, child-rearing emphases, availability of role models, and interpersonal factors may explain these differences.

Cultural attitudes toward aggression and violence are reflected in a society's actual levels of aggression and in the socialization of children. Cross-cultural studies have shown that some societies are more aggressive than others. Classic examples of aggressive societies are the Ik of Uganda (Turnbull, 1972), the Mundugumor of New Guinea (Mead, 1935), and Amazon populations such as the Yanomamo (Chagnon, 1968) and Waorani (Robarchek & Robarchek, 1992). In these societies, homicide rates are high and in-group conflict and out-group warfare are common. Children are taught to be independent, combative, and emotionally unresponsive to the needs of others, and aggression is a way of life. By some standards, the United States is also an aggressive society. On a percentage basis, its crime rates are higher than other industrialized countries (Wolff, Rutten, & Bayer, 1992), and its policies on gun control contribute to the frequency of inner-city gang warfare, armed robbery, and other violent crime.

Some research suggests that cultural differences in aggressive and prosocial behavior can be attributed to the emphasis a particular society places on competition and stresses individual rather than group goals. For example, in collectivistic (Triandis, 1990) societies like Mexico (Kagan & Madsen, 1971) and Korea (Kim & Choi, 1994), children are taught to suppress individualism, to cooperate with others, and to avoid interpersonal conflicts. On the other hand, in individualistic (Triandis, 1990) societies like the United States, parents often condone aggression by urging their children to stand up for themselves and to retaliate against aggression. This notion is supported by research from the relatively collectivistic culture of the Marquesas Islands. Martini (1994) reports that Marquesan children are sensitive to status differences and are intolerant of children who are competitive, try to boss others, or make rules for play. Instead, children maintain group solidarity and avoid social conflict and the need for play negotiations by engaging in familiar games that are played in the same way from one time to the next and where everyone performs the same actions at the same time.

Farver and her colleagues reported similar findings in a series of studies comparing the play behavior of Korean American and Anglo-American preschoolers (Farver & Kim, 1994; Farver, Kim, & Lee, 1995; Farver & Lee, 1997). They found that Korean American children, whose culture emphasizes a relational mode characterized by group interdependence and sensitivity to others, responded in a cooperative fashion to peers' play initiations and were nonconfrontational in their dyadic play. They avoided using communicative strategies that required them to direct another child's behavior, to set and enforce rules, or to decide roles or scripts. On the other hand, the Anglo-American children, whose culture emphasizes an aggregate mode, characterized by independence and a preoccupation with the self and its expression, were frequently aggressive and responded negatively to peers' initiations, rejected their partners' contributions, and their play was often conflictual.

Child-rearing practices also have implications for aggression, particularly with regard to whether parents punish or condone aggressive behavior. Some studies suggest that in the act of physically punishing their children for aggressive behavior, parents may be inadvertently teaching them to behave aggressively (Patterson, DeBaryshe, & Ramsey, 1989). Children who learn that they will be hit, kicked, or shoved when displeasing their parents will often direct the same kinds of responses toward playmates who displease them. Fry (1988) found comparable results in a study examining children's aggression in two Mexican Indian Zapotec towns. He found that the children who lived in the town with the violent reputation (i.e., high in societal and domestic violence and frequent physical punishment of children), performed twice as many aggressive acts as the children from the controlled violence town. Moreover, parents in the violence-prone town directly encouraged their children to be aggressive and often did not stop fights among children.

Studies have also shown that cultural practices encourage particular modes of reasoning about conflicts. For example, Kornadt (1990) administered an Aggression Thematic Apperception test to adolescents in Germany, Switzerland, Japan, Bali, and Batak to examine the relationship between aggressiveness and maternal style. His results showed that the German and Swiss children, whose mothers reacted to their children's misbehavior with frustration and anger, scored two times higher in levels of aggression than the Japanese, Balinese, and Batak children, whose mothers responded by exonerating or conceding to their children. The Western children also scored higher than the Eastern children on measures of inhibition of aggression, indicating that the Western children had more aggressive responses overall, and attempted at times, to repress their responses.

The opportunity to emulate aggressive models also varies across cultures. In societies where adults and older children seldom behave aggressively, young children will have few opportunities to learn aggressive behavior. In a study of the West Malaysian Semai, Denton (1978) reported that the most influential inhibitor of aggression is the fact that children see so few examples of it. In Semai society, adults rarely boss each other around or use physical punishment with children, "therefore if a child wanted to become violent, it would have no clear idea of how to proceed" (p. 132).

In sum, a child's tendencies toward aggression and antisocial conduct will depend, in part, on the extent to which the wider culture encourages and condones such behavior. However, not all children raised in an aggressive society are prone to aggression. Individual differences in children's aggression within cultures may be a consequence of variations in socialization histories. Specifically, researchers have found that aggressive behavior is positively related to neglectful child-rearing practices. Highly aggressive children tend to be raised in homes where parents are rejecting, lacking in warmth and affection, are indifferent or permissive of their child's expression of aggression, and who discipline their children using physical punishment (Eron, 1982; Olweus, 1980; Parke & Slaby, 1983).

Children's aggressive behavior has also been related to poor social competence with peers. Being socially skilled involves being able to initiate peer interaction, maintain ongoing interactions, and resolve interpersonal conflicts (Asher, Renshaw, & Hymel, 1982). Children who lack social competence often behave negatively with peers (Coie & Kupersmidt, 1981; Dodge, 1981; Newcomb, Bukowski, & Pattee, 1993), experience difficulty in initiating and maintaining interactions, and may use aggression to resolve conflicts (Dodge, Petit, McClaskey, & Brown, 1986).

There also appear to be gender differences in children's aggression. Although aggression is a reasonably stable trait for both genders (Caspi, Elder, & Bem, 1987), theorists do not agree on the extent to which there are gender differences in aggression. Early studies consistently found males to be more aggressive than females (Ember, 1981; Maccoby & Jacklin, 1974). However, recently, Bjorkquist (1994) and others (Hyde, 1984) have argued that gender differences in aggression are stylistic variations rather than quantitative differences. On the other hand, cross-cultural studies suggest that gender differences in aggression are related to the gender-typing practices of a particular society (Barry, Josephson, Lauer, & Marshall, 1976; Frey & Hoppe-Graff, 1994). That is, when cultural …