The Importance of Early Attachments

Neither Erikson nor Bowlby saw the positive outlook of the secure child as immutable. Whereas secure children play a role in making their positive expectations come true by their choice of partners and through their own responses to others (Sroufe, 1996), they are not all powerful. If later experiences, especially with caregivers, violate their expectations, their burgeoning ideas about self and others could be modified, incorporating more negative expectations. In sum, early attachment is said to launch processes that can have long-term consequences; but the quality of care that the child continues to receive can either strengthen or redirect those processes.

Does early attachment quality predict later psychosocial functioning? Sroufe, Egeland, and their colleagues followed a large sample of Minneapolis children, beginning with two assessments of attachment status in the strange situation at 12 and 18 months (e.g., Vaughn et al., 1979). Children whose attachment status was stable from the first to the second assessment were later evaluated on a variety of interpersonal and cognitive dimensions, and differences were found between the insecurely and securely attached children on many measures. For example, when the children were 4 years old, 40 of them participated in a summer nursery school program at the University of Minnesota (Sroufe, Fox, & Pancake, 1983). Teachers, who knew nothing about the earlier attachment ratings, ranked the children each day on characteristics related to autonomy, such as attention seeking, extreme reliance on the teacher, involvement with teachers at the expense of peers, and so on. The children who had been securely attached as toddlers were more often seen by teachers as direct and appropriate in their dependency behaviors, seeking help when they realistically needed it but functioning independently in other situations. Insecurely attached children were more likely to act helpless, to act out for attention, or in some cases to passively avoid seeking help when they genuinely needed it.

In assessments at later ages, the children in the Minneapolis project showed other continuities as well. In a summer camp at age 10, for example, securely attached children tended to be more self-confident, to have more friends, to have better social skills, and so on. Many studies, though not all, have found similar continuities (see Thompson, 1998, for a review of these findings).

If infant–mother attachments predict later developments, what about infant–father attachments? When infants are securely attached to both parents they seem to have the best outcomes; if they are securely attached to one parent, but not the other, their outcomes are better than if they are insecurely attached to both (e.g., Belsky, Garduque, & Hrncir, 1984; Main, Kaplan, & Cassidy, 1985; Suess, Grossmann, & Sroufe, 1992). Thus, a secure attachment relationship with a second parent adds a protective factor, perhaps helping to safeguard a child from the negative influence of an insecure attachment. Several investigators have also suggested that because fathers and mothers tend to interact differently with their infants, the attachment relationship with each parent may affect different aspects of the child’s psychosocial development. For example, in American families, mothers spend more time with infants in quiet, ritualized routines such as peek-a-boo games, where synchrony of eye contact, vocalizations, and facial expressions are key factors. Fathers tend to engage in more energetic, stimulating games (such as whirling the baby through the air); they are more unpredictable and exciting (MacDonald & Parke, 1986; Pecheux & Labrell, 1994). Of course, though there tend to be differences in the types of interactions infants have with each parent, there is much overlap, especially when fathers are highly involved in infant care. Whether infant attachments to mothers versus fathers affect different aspects of later social functioning is yet to be determined. One study is suggestive, however. Pecheux and Labrell (1994) found that 18-month-olds were more playful and more inclined to smile with a stranger if their fathers were present when the new person approached than if their mothers were present.

Overall, there is little doubt that infant attachment relationships are predictive of some later behaviors. What is presently uncertain is, how important are early attachments in producing such outcome differences? More specifically, does the early attachment actually have long-term effects, or could the more recent and current caregiving environment be responsible for the apparent continuities? In the Minneapolis study, for example, some children who had been securely attached as infants were having behavior problems by age 4. The quality of parenting during the preschool period helped account for this change. Mothers of these children were less supportive when engaging their children in educational tasks at age 3. They provided less encouragement and were less effective teachers than mothers whose children did not develop later behavior problems. Also, children who had been insecurely attached as infants but who were well adjusted at 4 years tended to have mothers who were supportive and effective in their interactions with their preschoolers (Erickson, Sroufe, & Egeland, 1985). It seems that the role of early attachments in later development can be diminished, and perhaps even eliminated, when parental acceptance, support, and responsiveness changes substantially, either rising to the challenges of the child’s new developmental stage or failing to do so (see also Easterbrooks & Goldberg, 1990; Youngblade & Belsky, 1992).

What, then, is the value of getting a “good start” with a secure infant attachment? It may be that the primary influence of infant attachments is their tendency to perpetuate themselves. Once a relationship pattern between parent and child is established, it tends to be repeated; when it is, the older child’s behavior is consistent with predictions from the early attachment. In his first relationships, a baby may well form an incipient “working model” of what to expect from interactions. That model affects his behavior and expectations in the future, but the model is in progress and will be reworked in the context of new interactions.