Sex Differences in Attachment Emerge in Middle Childhood: An Evolutionary Hypothesis

J. Bowlby’s (1969 ⁄1982) theory of attachment, focused as it was on the survival function of attachment behaviors that the ethology of the time emphasized, led to the expectation that there would be no sex differences in patterns of attachment. Modern evolutionary thinking, however, building on insights of life history theory, parental investment theory, and sexual selection, yields an alternative prediction—that adaptive sex differences in attachment should emerge in middle childhood and be present in adults, consistent with sexual differentiation of reproductive strategies. This article reviews the theoretical basis of this expectation, including the recent proposal that a hormonally driven reorganization of attachment occurs at the beginning of middle childhood. Available data and various methodological issues involved in empirically testing the proposed model are discussed. It argues that life history theory offers a powerful organizing principle for understanding the emergence of individual differences, providing developmental researchers with exciting opportunities for empirical discovery and theoretical synthesis. KEYWORDS—attachment; evolution; sex differences; middle childhood; romantic attachment Evolutionary biology lies deep in the DNA of attachment theory. Based as it was on John Bowlby’s brilliant integration of the psychology and ethology of his time (Bowlby, 1969 ⁄1982, 1973, 1980), attachment theory revolutionized the study of parent–child relationships and has served ever since as a Correspondence concerning this article should be addressed to Marco Del Giudice, Center for Cognitive Science, Department of Psychology, University of Turin, Via Po 14, 10123 Turin, Italy; e-mail: [email protected]. a 2010, Copyright the Author(s) Journal Compilation a 2010, Society for Research in Child Development Volume 4, Number foundation for much developmental thinking (see Cassidy & Shaver, 2008). The evolutionary theory that proved so central to Bowlby’s insights, however, exerted little impact on developmental psychology until the 1990s, when evolutionary models of reproductive strategies and attachment patterns advanced by Belsky, Steinberg, and Draper (1991), Belsky (1997a, 1999), and Chisholm (1993, 1996, 1999) recast attachment theory in modern evolutionary perspective (see Simpson & Belsky, 2008). By incorporating insights from life history theory (see below), Belsky et al. (1991) and Chisholm (1993) argued that children use attachment security in the first years of life as a cue of local ecological risk and uncertainty, thereby influencing the development of reproductive strategies. Insecure children, especially insecure-avoidant ones (see below), were expected to follow a trajectory of early reproduction and physical maturation; short-term, uncommitted relationships with partners; low parental investment; and increased opportunism and risk taking. Secure children, in contrast, would follow reproductive strategies involving later reproduction, longer term couple relationships, higher parental investment, and a more trusting, mutually beneficial approach to close interpersonal relationships. To date, researchers have provided empirical support for many predictions derived from this evolutionary framework (reviewed in Del Giudice, 2009a; Simpson & Belsky, 2008). At the same time, there is growing consensus that both parent–infant bonds and long-term couple relationships in adults involve the activity of the attachment motivational system. The evidence for this proposition comes from two main sources. First, the behavioral and psychological dynamics of adult bond formation, separation, and loss show striking similarities with those observed in children (Feeney, 1999; Hazan & Zeifman, 1999). Second, neurobiological studies have revealed substantial overlap in the neurochemical and neuroanatomical substrates involved in both types of relationships (e.g., Carter, 1998; Insel & Young, 2001; Pedersen et al., 2005). Of course, couple relationships are not identical to parent–child attachments, and