Childhood enemies

Noel A. Card and Deborah M. Casper ask whether they are necessary challenges or markers of social maladjustment

Around one in three children are involved in an enemy relationship at any given time, suggesting that they are a common experience. Yet it has not been determined if being involved in an enemy relationship should be considered normative or a marker of maladjustment. It is possible that enemies provide necessary developmental challenges, but it is equally plausible that they cause unnecessary distress. It is also possible that different types of enemies lead to differential outcomes. The research is in its early stages; this article reviews what is known and raises several important questions.

Enemies. The very word is both familiar and potentially uncomfortable for many people. Most adults can recall at least one individual they have considered, or possibly still do consider, an enemy (Card, 2007; Holt, 1989; Wiseman & Duck, 1995), many of which occurred during childhood or adolescence (Gürog?lu et al., 2009; Hartup, 2003). Yet despite our personal familiarity with these relationships, and the seemingly high prevalence of them prior to adulthood, scientific investigation of their role in development has only recently begun.

This article is motivated by a recent meta-analysis conducted by Card (2010) that summarised the research findings on the topic of child and adolescent antipathetic relationships (defined below). This meta-analysis included 26 studies, most of which were conducted in the last decade. Given the emergent state of this area of research, we want to explore, in this article, whether antipathetic relationships during childhood and adolescence should be considered normative challenges or areas for concern. Specifically, we organise this article around three questions. First, how common are these relationships during childhood and adolescence? Second, are these relationships associated with maladjustment? Third, how do these relationships vary in terms of their histories, characteristics, and associations with maladjustment? Following our consideration of these three questions, we will describe what we believe are the most important next steps in scientific understanding of the role of enemies in child and adolescent development.

Before turning to these issues, it is important to briefly clarify some of the terminology in this area of study. As is common in relatively new areas of study, some confusion related to terminology and definitions has emerged. Many have used terms such as ‘enemies’ or ‘enemy relationships’ to describe relationships based on mutual dislike (see Hodges & Card, 2003), but others argue that the term ‘enemy’ should be reserved for relationships marked by especially intense feelings such as hatred (Hartup, 2003). In this article, we rely on the terminology proposed by Card (2007), in which antipathetic relationship refers to a dyadic relationship defined by mutual dislike, and the term antipathy refers to the partner in that relationship. Although this terminology is less recognisable than terms such as ‘enemy’, this terminology is explicit and consistent with common measurement practices in the field.

In considering the potential importance of studying antipathetic relationships, an important initial question is how common they are. Estimates of the prevalence rates of antipathetic relationships during childhood and adolescence vary tremendously. Perhaps the single best prevalence estimate comes from a recent meta-analytic synthesis of prevalence rates in 16 studies (Card, 2010), which estimated an average prevalence rate of 35 per cent. In other words, about 1 in 3 children and adolescents have at least one antipathetic relationship at any given time, averaged across the available studies.

To investigate the sources of differences across studies, the meta-analysis used coded study characteristics as predictors of study prevalence estimates. Results indicated that the differences could be explained in part by study methodologies, but these methodological features alone did not account for all of the variability in prevalence rates across studies. Card (2010) found a very small effect size (r = .05) tendency for boys to have more antipathetic relationships than girls. Investigations of potential ethnic differences in prevalence are limited, but two studies (Abecassis et al., 2002; Hembree & Vandell, 2000) have suggested the absence of substantial differences. From a developmental perspective, an important question is whether prevalence of antipathetic relationships changes with age. In the meta-analysis, the mean age of study samples was not associated with prevalence rates, suggesting an absence of pronounced patterns of antipathetic relationships becoming more or less common across child and adolescent development.

In sum, the available evidence indicates that about one in three children and adolescents are engaged in an antipathetic relationship at any given time, and the evidence thus far does not indicate substantial differences in prevalence by gender, ethnicity or age (at least among school-age children and adolescents).

In contrast to the minority (albeit a substantial minority) of children and adolescents who have an antipathetic relationship at any given point, a large majority (i.e. about 70–75 per cent) of young adults (i.e. college students) can retrospectively recall having an antipathetic relationships at some point (Card, 2007; Casper & Card, 2010; Holt, 1989). Therefore, it appears that most individuals experience these relationships at some point, if not continuously.

Associations with maladjustment
A second critical question to ask when considering the importance of studying antipathetic relationships is whether they are associated with maladjustment. Early writings on the topic varied. One view was that antipathetic relationships are related to maladjustment because they negatively impact those engaged in them (e.g. Card et al., 2001, described the ‘dangers of enemies’). A second view was that individuals ‘need’ antipathetic relationships, suggesting that these relationships may be associated with positive adjustment, or at least not associated with maladjustment (e.g. Abecassis, 2003; Barash, 1994).

The research, however, paints a more worrying picture. Card (2010) summarised the available concurrent associations between the number of antipathetic relationships children and adolescents have with several aspects of psychological, academic and social maladjustment. These results indicated a clear pattern of associations with maladjustment. Antipathetic relationships are associated with more externalising problems, such as aggression and oppositional behaviours, as well as internalising problems, such as depression and social withdrawal. Antipathetic relationships are also associated with lower levels of adjustment, such as enactment of prosocial behaviours and academic performance. In terms of social adjustment, antipathetic relationships are associated with other social problems in the form of higher victimisation and rejection by peers, less positive peer regard (e.g. being liked by peers), and fewer friendships. This evidence clearly points to antipathetic relationships being associated with maladjustment.

There are several limitations to these findings, however. First, Card (2010) considered only concurrent (i.e. single time point) correlates, which do not speak to whether antipathetic relationships cause maladjustment or maladjusted individuals are more likely to form antipathetic relationships (or both). There exist only a few longitudinal studies that can shed light on directions of influence, and their findings are mixed. Considering a wide range of adjustment indices, both Abecassis (1999) and Pope (2003) found little consistent evidence that antipathetic relationships during middle childhood and early adolescence predict later maladjustment indices (four years and one year later, respectively). Murray-Close & Crick (2006) found inconsistent evidence that antipathetic relationships among 8- to 10-year-olds predicted overt and relational aggression and victimisation six months later. Another study investigating slightly younger children (second and third graders; approximately 7–8 years old) found that antipathetic relationships predicted subsequent (overt) aggression among boys but not girls (Erath et al., 2009).

In sum, the very limited longitudinal evidence is unclear about whether antipathetic relationships predict subsequent maladjustment. This directionality is plausible, given that having antipathetic relationships might be a stressful experience and might lead to negative interactions (e.g. victimisation by the antipathy: Card & Hodges, 2007). However, it might also be the case that some adjustment problems contribute to the formation of antipathetic relationships; for example, aggressive, immature and socially unskilled behaviour might lead to the formation of antipathetic relationships and the maintenance of existing ones. These possibilities are not mutually exclusive.

A second limitation to the findings of Card (2010) is that the associations between antipathetic relationships and maladjustment are consistently small in magnitude. One interpretation of these small effect sizes is that the associations are truly weak; that maladjustment plays a small role in the formation of antipathetic relationships and/or that antipathetic relationships have only a small contribution to maladjustment. However, we argue that a second interpretation should be considered: that antipathetic relationships are diverse, and therefore likely to have different risk factors and consequences for depending on the type of relationship and the role that child has in the relationship. For instance, if an antipathetic relationship promotes one child to aggress against the antipathy, then the outcomes differ for two children engaged in this relationship and thus the average association with maladjustment is necessarily weak. We turn attention to other potential aspects of the diversity of antipathetic relationships next.

Variation across relationships
The third question to answer in considering the importance of studying antipathetic relationships is whether (and how) these relationships vary in terms of their characteristics, histories and associations with maladjustment. If there is evidence of variation in these relationships, and this variation predicts maladjustment of the individuals involved, then a focus on this variation both (a) accounts for the modest overall associations between antipathetic relationships and maladjustment, and (b) points to the identification of sources of these variations as an important line of future research. Although sparse, researchers have begun to provide preliminary evidence of diversity across antipathetic relationships based on individual characteristics, group characteristics, varying histories which likely involve different features (e.g. dissimilarity, jealousy, aggression, avoidance, reconciliation, etc.) and dynamic trajectories across the different stages of antipathetic relationships (e.g. formation, maintenance and termination).

One aspect of variation in antipathetic relationships is whether the partners are similar or dissimilar. This similarity and dissimilarity could be in terms of demographic characteristics (e.g. same- versus mixed-gender: Abecassis et al., 2002; Card, 2010), interpersonal behaviours (e.g. being withdrawn, antisocial or prosocial: Gürog?lu et al., 2009; Hayes et al., 1980; Hayes et al., 1996) and group membership (e.g. group antipathies based on membership in different social crowds: Laursen et al., 2010). It is plausible, though not yet tested, that antipathetic relationships marked by similarity versus dissimilarity may differ in their interactions and impact. For instance, it is possible that antipathies who are similar to one another are likely to occupy similar social niches (e.g. same peer groups or crowds), which in turn could lead to either partial reconciliations or greater impact due to more regular contact. In contrast, it is plausible that antipathetic relationships in which dissimilarities exist are more harmful, given that these dissimilarities can create a differential in power that can be exploited through aggression (there is evidence that attempts to harm the partner are more frequent in antipathetic relationships than in acquaintanceships or friendships: Card, 2007).

Antipathetic relationships may also differ in terms of their relationship histories. There is a small amount of evidence that a substantial percentage of antipathies are former friends. One recent study has shown that antipathetic relationships among former friends, who often share a long and rich history, involve features such as intimacy rule violations (e.g. divulging confidential information), but also efforts at reconciliation (Casper & Card, 2010). In contrast, antipathetic relationships among children who are not formerly friends may instead arise from rivalry or competition, and therefore might involve different features, such as annoyance and frustration. It is also possible that these two types of antipathetic relationships (i.e. former friends versus nonfriends) will involve similar features but at different levels of intensity with potentially different consequences. The different level of intensity is evident in the words of two adolescent females. The first describes her experience with an antipathy who was a former friend:

This female insisted and still insists that I am trying to interfere with her family because I am her brother's girlfriend. We were close friends before she began to think this. She introduced me to her brother and we are still together, but are having problems because of her. We want a future together, but I'm not so sure now.

The second describes her experience with an antipathy who was identified as a rival (but not former friend):

She would always want things her way and if it wasn’t done her way she would get mad cause sometimes we would chill [with the] same group of people and it was always the same problem. (Casper & Card, 2010)

These examples speak to how different features of antipathetic relationships can arise from the history of the relationships.

Unfortunately, very little work has been done to examine the features and course of different types of antipathetic relationships. Two studies have attempted to look at this, however. In comparison to friendships, Card (2007) found conflict to be a common feature in the formation of antipathetic relationships and aggression and avoidance to be common during the maintenance of the relationship. In another study comparing antipathetic relationships that formed from broken friendships to those that did not, Casper and Card (2010) found that antipathetic relationships between former friends were more likely to involve intimacy rule violations in their formation, relational aggression, competition, and avoidance during their maintenance, and continued contact after high school (i.e. they are likely to have mutual friends and will encounter one another more than if they were not formerly friends; it is also plausible that they will reconcile). Future work will need to continue this line of inquiry so as to better describe this diversity and then determine whether the various relationship types are differentially associated with maladjustment.

Next steps for research
As should be clear from our review of the existing literature, there is certainly much that we do not yet know about antipathetic relationships. Therefore, this area of investigation is ripe for further studies. In our view, the most pressing next steps for research lie at the limits of our knowledge around the three questions organising this article.

Regarding prevalence, we still cannot fully explain the widely varying estimates obtained across studies. Although methodological aspects of the studies play some role (Card, 2010), research dedicated to identifying variation in prevalence is valuable. Larger studies sampling a wider age range are needed to more clearly identify potential developmental trends (keep in mind that the Card, 2010, meta-analysis investigated average age within study as a rather imprecise predictor). Further investigation of demographic (ethnicity, socio-economic status) predictors will also be valuable. Perhaps most importantly, it will be useful to examine whether antipathetic relationships are more common in some contexts than others. For instance, multilevel studies of whether prevalence rates vary across school – and the predictors of this variation – would be useful in advancing our understanding of contexts that promote versus discourage antipathetic relationships.

Earlier, we noted that further longitudinal studies are needed to understand whether the maladjustment correlates synthesised by Card (2010) serve as antecedents or consequences, or have a bidirectional relation to antipathetic relationships. Such studies should consider a variety of time spans, considering both the short-term (i.e. days, weeks) and long-term (i.e. years) influences. Intensive longitudinal designs will be useful in identifying maladjustment associated with various stages (i.e., formation, maintenance, termination) of antipathetic relationships.

Finally, we described some ways that antipathetic relationships may vary, but at this point these possibilities are almost entirely speculative. Future research identifying the dimensions on which antipathetic relationships can be reliably distinguished, and showing the value of these distinguishing on these dimensions, are needed (e.g. differential associations with maladjustment). Once features of these relationships that predict different aspects of maladjustment are identified, then it will be possible to investigate the antecedents of these antipathetic relationship features. It will also be necessary to identify the salient features of antipathetic relationships if interventions are desired.

We framed this article around the question of whether antipathetic relationships during childhood and adolescence are necessary challenges or problematic. Our review of the available research has provided a mixed answer to this question. The prevalence rates we have reviewed indicate that they are common, with about one in three children and adolescents having an antipathetic relationship at any given time, and most adults experiencing these relationships at some point. In this sense, then, they are apparently not especially problematic, as most of us have had these relationships and still fared well.

On the other hand, the studies reviewed by Card (2010) indicated consistent, albeit small, associations with maladjustment. Acknowledging the limitation that Card (2010) relied on concurrent associations, from which directions of influence cannot be inferred, this association appears to indicate that antipathetic relationships are markers of maladjustment. However, the existing longitudinal studies are limited and inconsistent, so we do not yet know whether any impact of antipathetic relationships is lasting versus short-lived. It may be the case that these relationships are only detrimental if they are long-lasting and/or repeatedly experienced (i.e. an individual who forms many antipathetic relationships over the course of development), though this possibility has not yet been tested.

Another possibility is that not all antipathetic relationships are harmful, but some are. For instance, being physically victimised by an antipathy is strongly associated with internalising problems and low self-worth (Card & Hodges, 2007). But perhaps more benign antipathetic relationships marked by nonaggressive rivalry are not. The existing research is virtually silent on these possibilities.

In sum, it is difficult to answer whether antipathetic relationships are harmful or not – or even useful – given the existing research. At a minimum it seems safe to conclude that they can be painful and dangerous in the sense that harm could occur. However, they also represent challenges that most individuals experience during the course of childhood and adolescence; indeed, not experiencing and learning to manage such relationships may place people in a vulnerable position if they only acquire them during adulthood. Despite the uncertainties of the danger or value of antipathetic relationships, however, it is clear that there is much more we need to understand about these relationships.

Noel A. Card is Associate Professor in Family Studies and Human Development at the University of Arizona

Deborah M. Casper is at the University of Arizona

Abecassis, M. (1999). I dislike you and you dislike me: Prevalence and developmental significance of mutual antipathies among preadolescents and adolescents. Unpublished doctoral dissertation, University of Minnesota, Minneapolis.
Abecassis, M. (2003). I hate you just the way you are: Exploring the formation, maintenance, and need for enemies. New Directions for Child and Adolescent Development, 102, 5–22.
Abecassis, M., Hartup, W.W., Haselager, G.J.T. et al. (2002). Mutual antipathies and their developmental significance. Child Development, 73, 1543–1556.
Barash, D.P. (1994). Beloved enemies: Our need for opponents. Amherst, NY: Prometheus Books.
Card, N.A. (2007). ‘I hated her guts!’: Emerging adults’ recollections of the formation, maintenance, and termination of antipathetic relationships during high school. Journal of Adolescent Research, 22, 32–57.
Card, N.A. (2010). Antipathetic relationships in child and adolescent development: A meta-analytic review and recommendations for an emerging area of study. Developmental Psychology, 46, 516–529.
Card, N.A. & Hodges, E.V.E. (2007). Victimization within mutually antipathetic peer relationships. Social Development, 16, 479–496.
Card, N.A., Hodges, E.V.E. & Isaacs, J. (2001, August). The danger of enemies: Relations with peer victimization. Paper presented at the New England Social Development Consortium, University of Connecticut.
Casper, D.M. & Card, N.A. (2010). ‘We were best friends, but…’: Antipathetic relationships emerging from broken friendships. Journal of Adolescent Research, 25, 499–526.
Erath, S.A., Pettit, G.S.,Dodge, K.A., Bates, J.E. (2009),  Who dislikes whom, and for whom does it matter: Predicting aggression in middle childhood, Social Development, 18(3),  577–596.
Güroglu, B., Haselager, G.J.T., van Lieshout, C.F.M. & Scholte, R.H.J. (2009). Antagonists in mutual antipathies. Journal of Research on Adolescence, 19, 35–46.
Hartup, W.W. (2003). Toward understanding mutual antipathies in childhood and adolescence. New Directions for Child and Adolescent Development, 102, 111–123.
Hayes, D.S., Gershman, E. & Bolin, L.J. (1980). Friends and enemies: Cognitive bases for preschool children’s unilateral and reciprocal relationships. Child Development, 51, 1276–1279.
Hayes, D.S., Gershman, E. & Holteman, W. (1996). Enmity in males at four developmental levels: Cognitive bases for disliking peers. Journal of Genetic Psychology, 157, 153–160.
Hembree, S.E. & Vandell, D.L. (1999). Reciprocity in rejection: The role of mutual antipathy and children’s adjustment. Unpublished manuscript.
Hodges, E.V.E. & Card, N.A. (Eds.) (2003). Enemies and the darker side of peer relations. New Directions for Child and Adolescent Development, 102. San Francisco: Jossey-Bass.
Holt, R.R. (1989). College students’ definitions and images of enemies. Journal of Social Issues, 45, 33–50.
Laursen, B., Bukowski, W.M., Nurmi, J. et al. (2010). Opposites detract: Middle school peer group antipathies. Journal of Experimental Child Psychology, 106, 240–256.
Murray-Close, D. & Crick, N.R. (2006). Mutual antipathy involvement: Gender and associations with aggression and victimization. School Psychology Review, 35, 472–492.
Pope, A.W. (2003). Developmental risk associated with mutual dislike in elementary school children. New Directions in Child and Adolescent Development, 102, 89–110.
Wiseman, J.P. & Duck, S. (1995). Having and managing enemies. In S. Duck & J.T. Wood (Eds.) Understanding Relationship Process Series: Vol. 5. (pp.43–72). Thousand Oaks, CA: Sage.

BPS Members can discuss this article

Already a member? Or Create an account

Not a member? Find out about becoming a member or subscriber