Attachment is an emotional bond, usually between child and parent, characterized by the child’s tendency to seek and maintain proximity to the parent, especially under stressful conditions. John Bowlby, a British psychoanalyst, developed attachment theory in the 1950s and 1960s as a way of explaining certain elements of personality and of psychopathology that were not already accounted for by other psychoanalytic theorists. Speciﬁcally, the genesis of the theory comes from Bowlby’s work with juvenile delinquents in pre-World War II England. He was impressed by how often the young criminals’ early experiences included severe disruptions in their relationships with their mothers. At the time, most psychoanalytic theory concerning childhood experiences was based on retrospective interviews with adults, and so Bowlby’s plan to study children via direct observation was nothing short of revolutionary. The result was an interesting and rather surprising blend of traditional Freudian psychoanalysis (see Sigmund Freud) and evolutionary theory.
Inﬂuenced by Konrad Lorenz’s ethological studies of imprinting among animals, Bowlby believed that humans might also be biologically predisposed to form a powerful, long-lasting bond to a speciﬁc individual. Such a bond would certainly serve a survival function, especially given the extreme helplessness of human infants, and so the idea ﬁt right in with an evolutionary perspective. Bowlby was also inﬂuenced by Harry Harlow’s work with rhesus monkeys. In a classic experiment, Harlow presented infant monkeys with a choice of two surrogate mothers, both made of chicken wire. One was made of bare wire but was equipped with a nipple that dispensed milk. The other provided no food, but had a soft covering and was warm. Harlow found that contact comfort (access to the warm, cloth-covered “mother”) was more important to infants than the food. This discovery upset the then-dominant view that the physiological need for food assured attachment.
Combined with the psychoanalytic belief in the power of early experiences to either produce lifelong psychological adjustment or lifelong psychological problems, the result of Bowlby’s evolutionary thinking was a theory in which failure to form a healthy mother-child attachment could serve as an explanatory mechanism for many developmental problems later in life. (As with so much of psychoanalytic thinking, mothers are central to the theory, whereas fathers barely rate a mention at all.) According to Bowlby, the mother-child bond is formed during a sensitive period in childhood and is carried forward through the rest of the lifespan in the unconscious mind.
Unlike most other theories rooted in psychoanalysis, however, attachment theory has proven fairly popular among more scientiﬁcally oriented developmental psychologists (although they tend to discard the notion of the unconscious inﬂuence on all other psychological development), largely because of its roots in actual observational research on children. The primary laboratory technique used to measure the attachment relationship is the “Strange Situation,” developed by Mary Ainsworth, which has now been used in thousands of studies. In the Strange Situation, a mother and child (or, in some recent research, father and child) are brought into a laboratory space, typically containing two chairs and a pile of toys in the middle of the ﬂoor. The room is usually equipped with a two-way mirror to allow for unobtrusive observation and/or videotaping. The only other participant is a stranger, usually a woman, for whom the primary selection criterion is that she must genuinely be completely unfamiliar to the baby. The Strange Situation typically consists of the following eight phases, though some variations have been occasionally used:
- Parent and infant are introduced to the experimental room.
- Parent and infant are alone. Parent places infant on the ﬂoor near the toys and sits in a chair. Parent does not participate while infant explores.
- Stranger enters, sits in chair, converses with parent. Stranger then approaches infant and engages in play with infant. Parent then leaves inconspicuously.
- First separation episode: stranger’s behavior is responsive to that of infant—if infant is comfortable with stranger, play continues. If infant is distressed, stranger attempts to soothe infant. The timing of this episode is quite variable—if the baby is highly distressed, the next phase begins immediately.
- First reunion episode: parent greets and comforts infant. Stranger leaves. Once infant is calm, parent leaves again.
- Second separation episode: Infant is alone.
- Continuation of second separation episode: stranger enters and gears behavior to that of infant. As with the ﬁrst separation episode, the second separation episode may be brought to an end prematurely if the infant is highly distressed—the goal is to make accurate observations of the parent and child, not to traumatize the baby.
- Second reunion episode: parent enters, greets infant, and picks up infant; stranger leaves inconspicuously.
The child’s behavior during the reunion episodes is the primary basis for classifying the infant’s attachment. The Strange Situation is usually used with children who are one year of age or older, for several reasons. First, mobility is necessary, and most children are at least crawling by that age, if not actually attempting to walk. Second, there are crucial developmental milestones, usually reached by that age, which are necessary for the attachment relationship to be evaluated.
For at least their ﬁrst half-year, many infants appear fairly undiscriminating in their affections, showing equal levels of comfort with most adults, whether friend or stranger. By the ﬁrst birthday, however, most have started showing signs of stranger anxiety (also called stranger distress), as well as separation distress (crying and general emotional upset in response to the departure of the mother or other primary caregiver). Both of these are among the signs sought as evidence of the quality of attachment in the Strange Situation, along with such things as proximity-seeking behavior and body language and eye contact upon reunion.
Based on infant behaviors in the Strange Situation, Ainsworth has identiﬁed three distinct patterns of attachment responses, resulting in three attachment classiﬁcations: Secure, Anxious/Avoidant, and Anxious/Resistant. Secure infants explore freely while the mother is in the room, making frequent eye contact and returning to the mother’s side from time to time—this behavior is frequently described as using the mother as a “home base” for exploration. These infants show some distress when left with the stranger, but they reunite with the mother enthusiastically and calm down very quickly. Anxious/avoidant infants sometimes show little distress when the mother leaves, and they actively avoid the mother when reunited. Anxious/resistant infants are distressed throughout the procedure.
Attachment classiﬁcations appear to result from the interaction of several variables, including both maternal responsiveness and infant temperament. Mothers who are sensitive to infant needs and adjust their behavior to that of their child have securely attached infants. Secure attachment has been found to impact positively other characteristics of the infant’s life as he or she grows up. Among other things, securely attached infants, unlike other infants, may grow up to be more curious and more comfortable with exploration of new situations, as well as better problem solvers. They also tend to be more socially competent and less likely to experience emotional problems.
Attachment has become a popular outcome variable in child development research, as the positive or negative impact of various childhood experiences on development is often assessed through attachment classiﬁcations. One of the more controversial uses of such data has been in the study of the impact of day care on child development. As more families switch from a single income to dual income, requiring more and more non-parental child care, developmental psychologists have become concerned over the impact this might have on the children’s social and emotional development. A large-scale review of research by Jay Belsky and David Eggebeen created a tempest when it came out in 1991, due to its conclusion that children who were placed in full-time day care during the ﬁrst six months of life were less likely to be securely attached to their mothers (though the article argued persuasively that the relative quality of care was a major confounding factor). The primary criticism of the use of attachment data in research on day care, and to a lesser extent in research on children of divorce, is that the American family, and the behavior society expects of it, has changed dramatically since Bowlby ﬁrst identiﬁed the importance of the attachment bond. Speciﬁcally, the behaviors sought as evidence of a secure attachment (moderate distress upon mother’s departure, stranger anxiety) may not be appropriate to expect in a child whose mother drops him off, sometimes to a stranger, on a daily basis. That child may appear insecure to some observers simply because he doesn’t react with distress.
Attachment has also been shown to vary according to different national and cultural contexts. Compared to American children, more German children are insecure/avoidant, more Japanese children are insecure/resistant, and Israeli children raised on a kibbutz (a communal, collective farming community in which children are not reared by their own parents) are more insecure/resistant. Within the United States, attachment classiﬁcations vary widely according to ethnicity. African American infants, who often have multiple caregivers, are less reactive than white children to the Strange Situation. Hispanic mothers intervene more with their children to maintain suitable public behaviors. Similarly to the international samples, this can lead to large numbers of children being classiﬁed as, respectively, insecure/avoidant or insecure/resistant. As with children in day care, these differences may reﬂect secure attachment consistent with cultural norms rather than suggesting unhealthy developmental outcomes.
Some researchers have attempted to extend research on parent-child attachment to adult populations, using a retrospective questionnaire called the Adult Attachment Interview. It classiﬁes adults’ childhood attachments into four categories (labeled differently than the established infant-attachment categories): autonomous, dismissing, preoccupied, and unresolved. This has allowed researchers to classify mothers’ childhood attachments and explore the relationship between their bond with their own parents and the caregiving style they adopt with their children. This research has found that maternal memories of and feelings about their own parents have an impact (either positive or negative) on the mothers’ own behaviors and beliefs as parents.
- Ainsworth, M. D. S., Blehar, M. C., Waters, E., and Wall, S. Patterns of Attachment: A Psychological Study of the Strange Situation. Hillsdale, NJ: Erlbaum, 1978;
- Belsky, J., and Eggebeen, D. Scientiﬁc Criticism and the Study of Early and Extensive Maternal Employment. Journal of Marriage & the Family, 53(4) (1991): 1107–1110.