Future Directions in…Friendship in Childhood and Early Adolescence

In 1935, the United States Congress proclaimed the first Sunday in August to be National Friendship Day. In 1997, no less than the United Nations named Winnie the Pooh as the world’s Ambassador of Friendship. Clearly, if the United States Congress and the United Nations have taken the time to discuss and vote on the significance of friendship, the phenomenon must be of some importance. If the reader is still not persuaded by the votes delivered by Congress or the U.N., perhaps the sage words of other significant individuals will prove convincing.

Friendship improves happiness and abates misery, by the doubling of our joy and the dividing of our grief.

Marcus Tullius Cicero, Roman statesman, orator, philosopher.

The proper office of a friend is to side with you when you are in the wrong. Nearly anybody will side with you when you are in the right.

Mark Twain, author

“And Pooh said to Piglet “Life is so much friendlier with two.””

A.A. Milne, author

Let us not forget the familiar stand-bys written by John Lennon (“I’ll get by with a little help from my friends.”) and Carol King (“You just call out my name, and you know wherever I am, I’ll come running to see you again.” from the song, You’ve Got a Friend).

Given the reputed significance of friendship, it should not be surprising that the social science community has long had an interest in this non-familial relationship. For example, in his oft-cited treatise on the interpersonal theory of psychiatry, Sullivan (1953) posited that the concepts of mutual respect, equality, and reciprocity developed from experiences with such “special” relationships as chumships, or friendships. Sullivan also wrote that with increasing age, peers, especially chums, aided each other in developing mature understandings of cooperation, competition and such social roles as deference and dominance. During preadolescence, concepts of equality, mutuality, and reciprocity became central to close friendships. Once acquired between friends, these concepts were thought to be extended to other relationships.

Contemporary researchers have drawn on Sullivan’s (1953) writings in citing the specific “services” that friendships contribute to the human condition. For example, friendships in later childhood and early adolescence are now thought to provide: affection, good company, and fun; emotional security; helpfulness, advice, and instrumental aid; validation of interests, hopes, and fears; opportunities for intimate disclosure; and prototypes for later romantic, marital, and parental relationships. Friends are thought to bolster feelings of self-worth and promote the growth of social skills and interpersonal sensitivity (e.g., Furman & Buhrmester, 1985; Rubin, Bukowski, & Parker, 2006). Perhaps the most important function of friendship is to offer children a pleasurable, extra-familial safe haven that reassures their exploration of the effects of their behaviors on themselves, their peers, and their environments.

The purpose of this brief essay is not to examine the voluminous extant and ever-growing literature on children’s and young adolescents’ friendships. Rather, we will focus on issues that require address should we wish to better understand all that which politicians, authors, musicians, and philosophers already appear to know so much about.

Some Very General Conclusions about the Friendships of Children and Young Adolescents

From the existing corpus of investigations, one can safely conclude the following about friendships in childhood and early adolescence (see Rubin et al., 2006 for an extensive review of this material). To begin with, there appear to be three operationally defining features of friendship: (1) Each member of the dyad affirms the existence of the friendship; (2) The relationship derives primarily from mutual affection; each partner views the other partner, and the relationship itself, as pleasant, fun, and likable. At the same time, the primary motivation for the relationship is not instrumental need; and (3) the relationship is voluntary; it is neither obligatory nor prescribed (Bukowski, Newcomb, & Hartup, 1996; Rubin et al., 2006). Significantly, regardless of age or sex of child/adolescent, most children and young adolescents have at least one same-sex mutually agreed upon friend. Typically, prevalence rates, regardless of child age, are between 60-to-80 percent, depending on whether the research focus is on “best” friendships (mutual nominations for best friend) or on any existing friendship (Howes, 1983; Parker & Asher, 1993; Rubin, Wojslawowicz, Rose-Krasnor, Booth-LaForce, & Burgess, 2006). Of course, it is important to note that some children without any mutual in-school friendships may have significant friendships with children who attend other schools, live in their neighborhoods, or meet with them regularly in the context of outside school activities. Despite these relatively high prevalence rates, friendship involvement is fluid; not all friendships (at least those that are school-based) are maintained throughout a given year. Yet, most children who “lose” a best friend are able to replace their “old” best friendship with a “new” one (Wojslawowicz Bowker, Rubin, Burgess, Booth-LaForce, & Rose-Krasnor, 2006).

Most friendships are formed between children who are, in many ways, racially, ethnically, and behaviorally similar to one another (Aboud & Mendelson, 1996; Haselager, Hartup, van Lieshout, & Riksen-Walraven, 1998; Vitaro, Tremblay, Kerr, Pagani, & Bukowski, 1997). Indeed, homophily appears to be a central, albeit not always beneficial, construct in children’s friendships (Berndt, 1996). For example, high academic achievers and those who are highly achievement-oriented are more likely than less academically-oriented children to have strong students as best friends; aggressive children are more likely than non-aggressive children to have other aggressive children as best friends; shy children are more likely than non-shy children to have other shy children as best friends; victims have best friends who are also victimized; girls (boys) are more likely to have girls (boys) than boys (girls) as best friends; and so on.

The significance of having and maintaining friendships is underscored by evidence suggesting that being friendless is harmful to a child’s psychological well-being and adjustment. Children who are without mutual friends, for instance, tend to report higher levels of loneliness and internalizing problems that those children with friends (Ladd & Troop-Gordon, 2003; Laursen, Bukowski, Aunola, & Nurmi, in press; Parker & Asher, 1993); chronic friendlessness appears to be the worst case scenario (Parker & Seal, 1996). Importantly however, gaining a friendship can significantly improve the lives of friendless children. When a previously friendless child gains a friend over the course of the school year, he or she is less likely to be victimized than are chronically friendless children (Wojslawowicz Bowker, et al., 2006).

Having a friend also seems to function as a protective factor for those children who are at risk for problems within the peer group. Children who are at risk for victimization because of their own personal proclivities (e.g., being aggressive or withdrawn) are less likely to experience victimization in the peer group if they are befriended rather than friendless (Hodges, Malone, & Perry, 1997). Whilst peer victimization predicts increases in internalizing and externalizing difficulties during the school year for those children without a mutual best friendship, it does not for children who do have a mutual best friendship (Hodges, Boivin, Vitaro, & Bukowski, 1999). This protective function also plays out for young adolescents whose relationships with their parents are less than optimal. For example, Rubin, Dwyer, Booth-LaForce, Kim, Burgess, & Rose-Krasnor (2004) found that when young adolescents reported difficulties in their relationships with their mothers and fathers, having a strong, supportive best friendship buffered them from negative self perceptions and internalizing problems.

All-of-the-above suggests that friendship is a significant and positive factor in the lives of children and adolescents. However, the world of children’s and young adolescents’ friendships is far more complex than typically portrayed in the peer relationships literature. Below, the reader will find a number of friendship-related questions and issues that, if addressed, could reveal new insights into the role that friendships play in the psychological, social, and emotional lives of children and adolescents.

1. Does Friendship Matter in the Lives of Children?

We have noted above that considerable evidence exists for the significance of friendship in children’s (and young adolescents’) lives. Specifically, researchers have shown that friendship may serve as a buffer or protective factor for some children who experience difficulties with peers in school and with parents in the family. We also know that individuals of all ages engage in more complicated social activity, talk, task orientation, cooperation, negotiation, prosocial (altruistic) activity, positive affect, and effective conflict management during social interactions with friends than non-friends (e.g., Simpkins & Parke, 2002). Friends are more responsive to each other than are non-friend dyads. And friends, as compared with non-friends, make more use of negotiation and disengagement, relative to standing firm, in their resolution of conflicts (e.g., Tomada, Schneider, & Fonzi, 2002). In terms of conflict outcomes, friends are more likely to have equitable resolutions (e.g., Newcomb & Bagwell, 1995). The differences described above suggest that children do view friendship as a unique context, separate and qualitatively different from their experiences with non-friends.

But can these findings be taken to mean that the experiences of friendship actually promote positive inter- and intrapersonal development, beyond those social, cognitive, and psychological benefits derived from interacting with parents and peers in general? For example, is there evidence to suggest that friendship per se promotes the development of the communicative and negotiation skills, understanding of compromise, and social consequential thinking demonstrated during observations of interactions between friends? Does friendship advance the development of empathy and sympathy? Does it promote respect and trust? Indeed, although positive affect is a central defining property of friendship, does friendship contribute to childhood happiness and positive psychological development? Despite the reputed positive correlates, outcomes, and significance that we have attributed to friendship, we know very little about such matters. We do know that there is empirical support for theoretical assumptions that peer interaction (specifically co-constructive exercise) is a causal force in social-cognitive and cognitive growth and development (e.g., Azmitia & Montgomery, 1993; Nelson & Aboud, 1985; Newcomb & Bagwell, 1995). This line of research clearly demonstrates that the interactions that occur more frequently among friends are more influential and facilitative of development than interactions occurring between non-friends (e.g, Miell & MacDonald, 2000). That is, a positive dyadic relationship may provide a better peer interaction learning context than a neutral or agonistic one.

Yet, many questions about positive development and friendship remain, and there is a need for longitudinal studies that involve children with friends and those without, and that permit the identification and comparisons of children who gain friendships and those who do not. As we go about addressing these questions, we would do well to consider whether friendship promotes different aspects of positive growth and development at different ages. What, in fact, are the provisions and products of friendship for a toddler, preschooler, elementary-, middle- and high-schooler? The challenges associated with developmental research should not deter us from uncovering its intricacies.

2. Does it Matter who the Child’s or Young Adolescent’s Friend Happens to Be?

Most children have at least one mutual friend, and many children have a group or network of close friends. As noted previously, evidence indicates that many children become friends with those who resemble them racially, behaviorally, emotionally, attitudinally, and developmentally (Aboud & Mendelson, 1996; Rubin, Lynch, Coplan, Rose-Krasnor, & Booth, 1994; Vitaro, et al., 1997). Of course, homophily with reference to central behavioral tendencies (e.g., aggression) or physically salient attributes (e.g. race, gender, or age) may characterize some friendships, but not others. Little is known, however about whether such similarity is actually beneficial to each member of the friendship or to the relationship itself. For example, it is heretofore unknown to what extent similarity functions to maintain the relationship and/or promote positive relationship qualities (for an exception, see Cillessen, Jiang, West, & Laszkowski, 2005).

Relatedly, does having a friend who demonstrates similarly maladjusted behaviors, (e.g. aggression, shyness) or who has similar peer difficulties (e.g. is victimized by peers) serve to amplify extant difficulties or do all friendships function protectively regardless of the partner’s characteristics? Although similarities in behavior and salient dispositions may serve as magnets of interpersonal attraction, we propose that the extent to which friendships are maintained may have more to do with need fulfillment. Put another way, with whom children become and remain friends may not simply be “another person like me.” Rather, it may well be “some person like me who likes me and who meets this need or that.” One can imagine, for example, that a shy/withdrawn child may befriend and continue a relationship with a similarly-behaved child who provides companionship and is comfortable doing so on the periphery of the social scene. Whether homophily with regard to central behavioral tendencies and observable physical characteristics is as important to friendship maintenance as are the provisions of opportunities to have fun and keep good company, self-disclose, offer support and security, offer advice, and so on is generally unknown. Also unknown is whether the provisions serving to maintain and enrich friendships (and development) vary with both age and the central behavioral tendencies of the friendship partners.

Moving beyond homophily, there has been a surprising dearth of research on the topic of friendship competence. Who are the children who become the “best” of best friend partners? Who are the children demonstrating supportiveness, responsiveness, trustworthiness, and positive affect in their friendships? Are these children equally effective across all of their friendship partners? Can a truly competent friend serve as an ameliorative agent for a less well-endowed partner? If so, at what ages can a competent partner help promote positive growth and development? In friendships within which one partner is clearly more competent (in a given domain) than the other partner, do both children benefit from the relationship? Given a Vygotskian perspective, might it be that only the “novice” or less skilled children benefit? In a similar vein, one might wonder about how friendships comprising an older and younger partner function. As individuals enter adolescence and high school, mixed-age friendships become increasingly normative, yet, there are virtually no studies on whether mixed-age friendships are advantageous in any way. Importantly, mixed-age friendships probably exist from early in childhood. Currently, we know very little about the same- and mixed-age friendships that children make and keep in their neighborhoods. Who are these neighborhood (out-of-school) friends? What roles do they play and what needs do they fulfill?

3. Friendship Provisions, Reciprocity, and Complementarity

Expanding on the notion of friendship needs, one may ask whether the search for friendship is based on ego-centered (Can you satisfy my needs?) or altruistic motivations (Do I have the ability to meet your needs)? Some children may seek a friend who can provide entrée into a particular social milieu. Others may seek friendship with a partner for whom they could offer fun, security, instrumental aid, and so on.

Once a friendship has begun, it has been assumed that they can be identified as more or less rich. “Richness” has been operationalized by the satisfaction of each partner, with regard to the provisions described by such theorists and researchers as Sullivan (1953), Weiss (1974), Selman (1980) and Furman (1996). Typically, “richness” has been defined by shared positive affect, supportiveness, felt security, intimacy, instrumental aid, and conflict resolution (see above). But is it the case that each member of a given friendship actually delivers each of these friendship provisions to the partner? Probably not. It is unlikely that partners engage in reciprocal exchanges with regard to each or any of the provisions of friendship; that is, an equal offering of provisions by each partner is unlikely. It is more likely that the “rule of richness” is one of complementarity rather than reciprocity; one partner offers more of X or Y whilst the other partner offers more of A and B. As long as the partners are satisfied with the extent to which the expected and required provisions are offered, and as long as the friendship is viewed as acceptable by each partner, the friendship is likely to be maintained. Have these notions been examined in the extant literature? We do not think so, but they should be!

In summary, it would seem timely to more carefully examine the intricacies of successful friendships. It is probably the case that as many friendships are based on complementary needs (e.g., Parsons, 1966) as on reciprocity and/or homophily. And it is equally likely that this latter statement is more or less true at different ages and for the two sexes.

4. Power and Control

The word “peer” typically connotes equality to another in abilities, qualifications, age, background, and social status. Piaget (1932) suggested that children’s relationships with peers could be distinguished, in both form and function, from their relationships with adults. The latter relationships were construed as being complementary, asymmetrical, and falling along a vertical plane of dominance and power assertion. Although it may be true that many friendships comprise partners who are “equal” or equivalent in age and sex, it is unlikely that the partners are equivalent in power and control (e.g., Savin-Williams, 1979, 1980; Updegraff et al., 2004).

What roles do power and control play in friendship maintenance and quality? Is it the case that one partner can wield more power in some domains whilst the other is more powerful in others? What are those domains? Many years ago, Parsons (1951) suggested that groups (e.g., families) typically comprise given individuals who take on the role of instrumental leadership whilst others take on the role of emotional leadership (Winton, 1995). Can friendships be characterized in this manner? Certainly, some provisions of friendships can be characterized as being more or less instrumental-oriented on the one hand, and nurturance-oriented on the other. Thus, in following from the third issue noted above, it may be that friendships are characterized as comprising partners who are equally powerful in all domains or differentially powerful in selected domains. This being the case, the definition of friendship quality may derive, in part, from the distribution of power and the roles played by each partner insofar as the delivery of provisions is concerned. Once again, these are issues that deserve future study within and across age and sex of friendship partners.

5. In the Eye of the Beholder

A friendship that offers most of the services of friendship noted above, and offers those services well, should aid both interpersonal and individual growth and development. Significantly, research that addresses the aforementioned statement is most often drawn from the perceptions of each individual friendship member. This raises the question of whether objectively appraised and observed friendship quality (e.g. Phillipsen, 1999) can capture the intrapersonal meaningfulness of the relationship to each partner. Put another way, it seems reasonable to ask: “What is the best index of friendship quality?” Is it subjective appraisal or the “objective” measurement drawn from (usually lab-based) observations? If a friendship is objectively observed to be insensitive, unresponsive, conflict-laden, emotionally vacuous, or interpersonally impoverished, and if the individual perceives the relationship as supportive, trusting, fun, and generally worthwhile, what are we left with? One may offer that such a scenario is beyond the realm of possibility, but it is an unanswered empirical question.

If one is interested in friendship longevity, an objective appraisal may prove more advantageous. If one is interested in the significance of friendship for self-worth and internalized thoughts and feelings of well-being, the personal perception may be advantageous. Importantly, however, the subjective appraisal of friendship quality as “good” when it is not may have negative consequences for both partners. A related matter is whether each partner views relationship quality in the same way. As suggested above, friendship quality may be defined differently given that each partner may have different relationship needs and expectations.

These issues may be of added significance if both partners view the friendship as “good” and the objective appraisal suggests that it is less than adaptive. In such cases, the relationship may become increasingly dangerous to each partner, as in an abusive or a dominance-submissive relationship. In our own research, for example, we have recently discovered that approximately seven percent of the best friendships of 5^th and 6^th grade young adolescents comprise an anxious/withdrawn and a volatile/aggressive individual; this is not an inconsequential percentage. We are now examining such matters as friendship stability and quality as well as whether these types of relationships represent risk factors for healthy development.

The bottom line is that we must begin to examine qualitative aspects of friendships both objectively as well as intrasubjectively, and in the case of the latter, it would make sense to capture the views of both partners – especially if one is interested in assessing relationship quality.

6. Sex Differences

Each of the questions and future research directions that we have noted thus far should be addressed from the perspective of the sex of the friendship partners. Some notable progress in understanding the role of gender in children’s friendships has been made (Rose & Rudolph, 2006). For example, it is well-known that girls’ friendships tend to be more intimate and caring than those of boys’ (Parker & Asher, 1993); boys tend to be more oriented toward the larger peer group whereas girls’ social orientations are directed to dyadic relationships (Maccoby, 1995).

With few exceptions, much of the sex differences research has been the product of self-reported questionnaire data. Might it be the case that boys’ think about and express their intimacy and care to their friends in ways not presently assessed by the extant batteries of friendship quality questionnaires? Perhaps too, some of the standard “friendship provisions” may be more important for girls than boys. Relatedly, some relationship qualities may be more or less important for boys and girls when determining the significance of the friendship for psychosocial well-being and development (Rose & Rudolph, 2006). To address these issues, observations of children and their friends are needed, along with queries (either in interview or questionnaire-format) into how boys and girls think about the forms and functions of their friendships. To this end, we need to evaluate the possible unique impact that specific friendship provisions might have on the formation, stability, and quality of boys’ and girls’ friendship. This too is an area of research that should be explored developmentally as the friendships of boys and girls may become more differentiated with age.

7. Race, Ethnicity and Friendship

As suggested above, children and adolescents tend to choose friends who are racially and ethnically similar to themselves. As a result, there are fewer cross-race than same-race friendships, and research demonstrates that the prevalence of cross-race friendships decreases with age (Aboud, Mendelson, & Purdy, 2003). Beyond the “homophily effect,” investigators argue that the decreasing number of cross-race peers available due to the homogenization of schools in the United States (Pettigrew, 2004),and strong in-group biases might account for the small number of cross-race friendships (Aboud et al., 2003). It is also possible that relatively homogenous subcultures within the United States, such as some religious groups and, for that matter, adolescent gangs, may actively discourage cross-race/ethnic or out-group friendships. Although differences in socioeconomic status are often controlled in studies of friendship, it may also be true that the complex relations between race, ethnicity, and SES account for the greater number of same-race/ethnic friendships than cross-race/ethnic friendships. Importantly, it has also been found that children believe cross-race friendships to be possible, even after a transgression, and that race is rarely used as the primary reason for why children become friends (Margie, Killen, Sinno, & McGlothlin, 2005).

The majority of the aforementioned research studies examined interactions occurring between and dyads comprised of same-race children only. Thus, it is unknown whether findings regarding the significance, processes, and provisions of friendship are generalizable to friendship dyads involving cross-race peers. There is some evidence indicating more intimacy in same-race than cross-race friendships, but the difference found was very small (Aboud et al., 2003); clearly additional research is needed.

It is possible that issues in cross-race/ethnic friendships may differ from those in same-race/ethnic friendships. For instance, similarities or differences in ethnic identity may be of more importance in cross-race friendships. Pettigrew (1998) has argued that the most beneficial way to change negative attitudes and perceptions is through interactions with individuals from different racial and ethnic groups. Taken together, it appears important for researchers to more carefully consider cross-race friendship involvement- not only because the provisions and processes of these relationships may differ from those of same-race friendships, but also because these relationships may play an important role in decreasing racial and ethnic discrimination.

8. Culture and Friendship

The three defining features of friendship listed above seem to be central for those who study friendship in Western, industrialized communities. It may be however, that these criteria are less than central in cultures within which mutuality, reciprocity, and voluntariness are viewed as less important by those who actually make decisions about children’s opportunities for friendship (e.g., Krappman, 1996).

The majority of psychological research on children’s friendships has involved middle-class Euro-American children and young adolescents. Within North America, little attention has been paid to culture (for exceptions, see Azmitia, Ittel, & Brenk, 2006; Graham & Cohen, 1997; Way, 2006). And beyond North America, surprisingly little peer relationships research has focused specifically on friendship.

This relative lack of cultural and cross-cultural research is rather surprising given that relationships within varying cultural communities are differentiated along such continua as individualism/autonomy, collectivism/connectedness, power distance, and uncertainty avoidance (Hofstede, 1980; Markus & Kitayama, 1991; Triandis, 1995), each of which has a relationship connotation. For example, researchers have noted that some cultures are relatively more collectivistic than individualistic in their values orientations. In these cultures, individuals are thought to form strong, long-term relationships with others and to value connectedness and conformity within those relationships as well as within the community at-large. One might surmise that in cultures leaning in the direction of collectivism, choice of friends is constrained by adult (parental) influence. It may be that extra-familial friendships remain less influential for longer periods of developmental time than relationships with parents and family members.

In contrast, in cultures that are relatively more individualistic, children and young adolescents may be freer to make their own choices of friendship, value independence, and more readily relinquish their extra-familial relationships when individual needs are not being met. And it may be that in such cultures, autonomy and extra-familial friendships assume more and earlier significance than is the case in more collectivistic cultures. Thus, it is possible that friendship involvement is most developmentally essential in cultures within which extra-familial peer relations are central. By-and-large, these “thoughts” about relationships, and specifically childhood and early adolescent friendships, are just that; little evidence exists to support these conjectures (for exceptions, see Azmitia et al., 1996; French, Bae, Pidada, & Lee, 2006).

Another, albeit understudied, culturally relevant construct is power distance. It is suggested that societies may be categorized by how much power specific individuals have within the larger community. For example, it may be that some cultures value relationships that are more egalitarian, whereas others may value a hierarchical relationship structure (Hofstede, 1980). One may ask whether friendship, when explored through the lens of culture, may reflect greater-or-lesser propensities in the directions of dominance/submissiveness and egalitarianism (see relevant discussion above).

Lastly, uncertainty avoidance involves the extent to which cultures feel comfortable in unstructured situations (Hofstede, 1980). According to Hofstede, uncertainty-avoiding cultures try to minimize the possibility of such situations by adhering to strict laws and rules, and on the philosophical and religious level by a belief in absolute truth. Consequently, one may expect that rules and regulations pertaining to relationships are clearly demarked through socialization practices. In such cultures, friendships may be selected by parents, not children. And the choice of friendship may be marked by perceived similarities in familial/cultural beliefs and traditions. These notions may be particularly valid for immigrant families (and especially parents) that aspire to cultural (and ethnic, religious) connectedness for their children. In cultures that are more accepting of uncertainty, there may be greater tolerance for philosophical and religious diversity. In this regard, there may be greater degrees of freedom accorded to both the nature of the friendship, who is considered an allowable friend (by parents and family), and how autonomous/independent the friendships (and the individuals) can be. Again, many of these notions have yet to be examined.

In summary, friendship researchers would do well to focus on culture, ethnicity, and race both within and across countries and communities. One good beginning would be to interview children and young adolescents about the very meanings of friendship (Bigelow & La Gaipa, 1975; Keller, Schuster, & Edelstein, 1993; Krappman, 1996; Selman, 1980). What is a friend and a friendship? What is it that defines a good friendship? How does one become a friend? How does one end a friendship? Thereafter, one could begin to observe the whos, hows, and whats of friendship, much of which was described and questioned above. Clearly, there is much to be done.