When the Children Cannot Yet Talk About Friendships

We define an affective relationship as one that includes feelings of affection or what would be called "love" in adult-child relationships. Toddler affective relationships have attributes of friendship common to the 'best friendships' which provide older children with emotional security and closeness (Howes, 1996; Howes, with Unger et al., 1992). These early friendship relationships appear to be formed in a way similar to adult-child attachment relationships (Howes, 1996). In the following section, we will examine supports for these assumptions about early friendship formation.

In a similar manner to the research on structural complexity of peer interaction, the friendship studies began with the collection of observational data. Since toddler-age children cannot report on their friendships, we must use behaviors to distinguish friendship relationships from playmate relationships in prelinguistic children. This results in some discontinuity in research about friendships because later research relies heavily on a child's ability to talk about friendships. For example, reciprocity of friendship is an important dimension of later friendship research, but cannot be explored in early childhood. Because early friendships can be defined only based on observed behavior, the earliest identified friendships must be reciprocal, with both partners engaging in defining behaviors.

We assume, in this work a three-part criteria for friendship: preference, recognition of the other as a social partner, and enjoyment (Howes, 1983, 1988b, 1996). Within this definition, children must prefer the company of the friend over the company of others and must enjoy the time they spend together to be considered friends. To operationalize these constructs for observational research, the following criteria were developed: (a) proximity (being within 3 feet of each other at least 30% of the observational period), (b) at least one instance of complementary and reciprocal play to indicate recognition of the other as a social partner, and (c) shared positive affect (both children expressing positive affect while engaged in interactive social play).

This definition was first used in a year-long longitudinal study of five peer groups (Howes, 1983). The children ranged in age from 10 months to 5 years. Each peer group was composed of same-age children. Eight times over the course of the year, we observed each child's interaction with every potential partner in the group. These observations were used to identify friend pairs. To test the assertion that toddler friendships were indeed affective relationships, we compared the three components of the behavioral definition (preference, recognition of the partner, and shared affect) to determine which component would identify the fewest friend pairs. The most stringent component was assumed to be most essential to friendship. Using shared positive affect to define friends identified the fewest friendships and, thus, was the most stringent part of the definition (Howes, 1983). This supports the premise that early friendships are affective relationships based on mutual affection. What remained to be explored, however, were the processes of friendship formation. The 1983 study included three groups of typically developing children in full time child care and two groups of children in therapeutic settings. The processes of friendship formation were somewhat different in the typical and atypical children, providing a glimpse of how this process occurs as well as the aspects that might be essential for typical social development.

Typical Children

As found in the first longitudinal descriptive study of infants and toddlers new to peer groups, the earliest friendships appeared when the children were 10 months old (Howes, 1983). Children in this early toddler period (10 to 24 months) tended to form only one or two friendship relationships. In three different longitudinal samples, these early toddler friendship relationships tended to be very stable, lasting a year in the first study (Howes, 1983). That is, once identified, friend pairs appeared at every subsequent observation period. In the two subsequent 3-year studies previously discussed, we again identified friend pairs at each observation period and found that friend pairs formed in the early toddler period reappeared for the 3 years of the longitudinal (Howes, 1988b; Howes & Phillipsen, 1992). That is, at each observation period, we identified friendship pairs within peer groups using observers blind to previous friendships. We then computed the probabilities of a particular friendship pair reappearing in more than one observation period. Friendship pairs formed when the children were toddlers were most likely to reappear in subsequent periods.

Of particular interest, given the prevalence of same-gender friendships in preschool and older children, friendships formed in the early toddler period are as likely to be same- as cross-gender in composition (Howes & Phillipsen, 1992). Further, cross- and same-gender friendships are equally likely to be maintained over time if formed in the toddler period (Howes & Phillipsen, 1992). This is in sharp contrast to the friendships of preschool- and school-age children who tend to form and maintain same-gender friendships.

Children in the late toddler period (24 to 36 months) had more friendship relationships than younger children, perhaps because they begin to differentiate between the different functions of friendship (Howes, 1987, 1988b). Some peers become friends, while others remain playmates. To be considered a playmate, but not a friend, a pair meets the criteria for preference and the recognition of the partner portions of the friendship definition but not the affective sharing component. In the late toddler period, friendships must still be identified by behaviors as children under 3 years of age cannot reliably complete a sociometric task. Again, the affective component of behaviorally defined friendship is critical in distinguishing friends from playmates (Howes, 1983).

By preschool age, children can play with children that they do not consider to be friends. Toddler-age children's social interactions are more fragile and more dependent on rituals and routines than the social interactions of preschoolers. Therefore, toddlers, more often than preschoolers, play only with their friends. We assume that this is because patterns of interaction between toddlers are highly ritualized.

Although behavior identification of friends is still a reliable measure in preschool (Howes, 1988b; Howes & Phillipsen, 1992), children also can reliably identify friends using sociometric ratings and nominations (Howes, 1988b). Beginning at age 3 and if they are enrolled in full-day child care and have a stable peer group, children select the same children in sociometric procedures as do observers (Howes, 1988b). This ability to communicate friendship status to another is a new skill. Perhaps this skill develops because preschoolers are now able to communicate the meaning of the construct of friendship. They use the language of friendship to control access to play ("I'll be your friend if you let me play").

While preschool age children are no longer dependent on rituals to sustain play, their play is still easily disrupted. For example, a pair of children may spend 10 minutes establishing the roles and scripts for a pretend play episode: "You be the lion and I'll be the little boy who finds you in the forest and then . . . " "No, I want to be a baby . . . " "OK, how about you be a baby lion and I'll be the little boy who finds you in the forest and . . . " "OK, and then when you take me home, you feed me with a bottle . . . " If a third child attempts to join, the play negotiations may have to start all over again and may not be successful. If the child attempting to enter the group is a friend, the other children appear more willing to undergo the negotiation process (Howes, 1988b). Children who are rejected by peers (using sociometric measures) but who have reciprocated friendships are more likely to be able to enter play groups because of having friends within the group (Howes, 1988b). Continuing the lion example, a child who is a friend might enter the play by saying, "Remember the time, I was the baby lion and then I got to be a great big lion and I roared but I didn't really hurt you. I'm going to be the daddy lion."

As discussed previously, just as there is stability in children's social interaction skills with peers (children who engage in complex play with peers in early developmental periods are the same children who are competent with peers in middle childhood), there is stability in children's friendship quality over developmental periods. This suggests an inter-relation between these two components of social competence—social interaction skills and friendship.

Atypical Children

Using observations rather than sociometric nominations to identify friends makes it possible to study friendship development in children who lack the social, cognitive, or linguistic skills to reliably complete sociometric interviews. We used the same behavioral identification of friends to study relationships among peers in two groups enrolled in outpatient programs for emotionally disturbed children directed by the child psychiatry unit of a mental hospital (Howes, 1983, 1984, 1985a). These children were diagnosed as severely emotionally disturbed with a predominance of nonorganic disorders. One group of children included toddlers, the other preschoolers. The children were observed eight times over the course of a year. Although the children in these groups formed fewer friendships than typical children, they did form friendships. That is, they demonstrated preference, recognition of the partner, and shared positive affect. However, unlike those of typical children, friendships among these atypical children were not sustained over time.

A second study examined friendship formation among abused toddlers who attended a daily child care program (Howes, 1984, 1988a; Howes & Eldredge, 1985; Howes & Espinosa, 1985). Again, these toddlers were found to form friendships as defined by preference, recognition of the partner, and shared positive affect. It is important to note that all of these children were enrolled in intervention programs which supported children's attempts to be prosocial and cooperative with peers. When abused children who were not enrolled in such programs were observed with their peers, a much different picture emerged. These children did not engage in shared positive affect, a necessary condition for friendship (Howes & Espinosa, 1985). These findings suggested that, in a context that supports positive relationships, very young children who have endured major disruptive life events can construct friendship relationships. Understanding which social contexts and which aspects of a particular social context can facilitate and enhance the development of trust and positive relationships among children who come from difficult life circumstances is an important area of future research.

Parenting Teens Special Report

Parenting Teens Special Report

Top Parenting Teenagers Tips. Everyone warns us about the terrible twos, but a toddler does not match the strife caused once children hit the terrible teens. Your precious children change from idolizing your every move to leaving you in the dust.

Get My Free Ebook

Post a comment