Pair Bonds as Attachments
Evaluating the Evidence
Anyone familiar with John Bowlby’s writings can readily understand the common (mis)perception that attachment theory applies exclusively to relationships between infants and their caregivers. Although he made repeated reference to attachment as a lifespan phenomenon, the principal focus of his theorizing was “the nature of the child’s tie to his mother” (Bowlby, 1958). His oft-cited claim that attachment is an integral part of human behavior from “the cradle to the grave” (Bowlby, 1979) was more a hypothesis than a thoroughly documented, empirically established fact.
The absence of a comprehensive theory of attachment beyond childhood may have delayed the initiation of research forays into the area, but it did not preclude them. In the last decade alone, investigations of adult attachment have proliferated at a rate comparable to that of infant attachment studies during the years immediately following the publication of Ainsworth’s (1967) original findings. Adult attachment research has proceeded largely on the faith that Bowlby was right about two things: that patterns of attachment established in early life are relatively stable across development, and that pair-bond relationships are the prototypical adult instantiation of attachment.
The present chapter does not address continuity in attachment patterns between infancy and adulthood. Instead, our focus is the second assumption—that romantic relationships qualify as attachment bonds and thus constitute the appropriate context in which to investigate adult attachment phenomena. Although these assumptions may appear to be inextricably interrelated, they actually represent independent issues, at least from an empirical standpoint. Consider the possible outcomes of stability studies: a finding of relative continuity in patterns of attachment from infancy to adulthood, or, alternatively, no systematic connections between infants’ strange situation classifications and their subsequent adult attachment categorization. Neither outcome would provide a definitive answer to the question of whether the attachment system is active in adult life or implicated in pair bonds. Continuity of individual differences is not the same as continuity of function; these are separate issues requiring distinct types of evidence. Thus the validity of our arguments concerning the second assumption is not dependent on the results of empirical investigations relating to the first.
The importance of the question—whether romantic bonds are attachments in the technical sense—can hardly be overestimated. The entire field of adult attachment research has been constructed on the premise that they are. If it were to turn out that Bowlby was mistaken, either about the lifespan significance of attachment or about