ATTACHMENT

History of Attachment Theory

The earliest roots of attachment theory can be found in Sigmund Freud's psychoanalytic theory of development, written at the turn of the twentieth century. Freud was the first theorist to propose a stage theory of development. His first stage, the oral stage, presupposed that infants develop relationships with their mothers, because mothers satisfy their hunger. Animal studies, however, provided persuasive evidence that feeding was not a sufficient explanation for attachment. In a series of famous experiments, Harry Harlow and his colleagues demonstrated that infant rhesus monkeys, raised in isolation, preferred the comfort of a cloth-covered surrogate mother to that of a wire-mesh surrogate with an attached feeding bottle. Clearly, the basis for attachment relationships does not reside in feeding alone. Erik Erikson, a student of Freud's, foreshadowed attachment theory by emphasizing the importance of children's ability to trust parents to meet their needs as the basis for later social and emotional development.

World Wars I and II alerted mental health professionals and the general public alike to the importance of close interpersonal relationships in development. Particularly in Europe, where casualty rates were highest, psychological trauma due to the loss of loved ones was common. Therapists, in fact, reported that death of family members was a frequent reason for individuals to seek therapy during the postwar years. In this context, British psychiatrist John Bowlby, while working with children and adolescents in London orphanages, discovered that the most disturbed children were those who had experienced separations from their caregivers, particularly their mothers. Consider, for example, his account of a seven-yearold girl:

At twelve months she fell ill with bronchitis and was in the hospital for nine months …During this time she never saw her parents, who were only permitted to visit her when she was asleep …When examined at the Clinic [at seven years of age] she was found …to be a withdrawn, detached, and unemotional child (Bowlby 1940, pp. 159-160).

Bowlby also noted that children who developed behavioral and emotional problems often experienced parenting that was characterized by displays of ambivalence or outright rejection. Based on these observations, he hypothesized that a caregiver's emotional attitude toward a child had direct implications for that child's later mental health. In other words, he believed that mental health is dependent upon a child feeling wanted and loved.

Three Main Propositions of Attachment Theory

Bowlby's seminal three-volume series on attachment and loss and subsequent work by his student, Mary Dinsmore Salter Ainsworth, form the core of attachment theory. There are three main propositions. The first is that infants' emotional ties to their caregivers can be viewed from an evolutionary perspective. Consider, for example, that closeness with adults can be viewed as an adaptive strategy for children because it leads to protection from environmental hazards, such as predators. Throughout the long evolution of human history, children who did not develop close relationships with their parents were less likely to survive and therefore less likely to reproduce. It is difficult to prove this thesis because there is no fossil record for social behavior. Still, it seems likely that attachment behaviors provided an evolutionary advantage.

Second, attachment is grounded in what is called a motivational control system, which organizes children's behavior. Just as physiological control systems are believed to regulate processes such as body temperature, a behavioral control system balances a child's desires to explore the environment and to seek proximity with caregivers, especially in the presence of danger. In this system, the child's primary goal is to feel safe and secure. Feelings of security, however, are dependent on caregivers' responses. When caregivers are sensitive and responsive, children are confidant that their needs will be met and that they may rely on their caregivers in times of stress. In contrast, when caregivers are insensitive and unresponsive, children become distrustful of their caregivers and are unable to rely on them. In the face of insensitive caregiving, infants develop strategies that are adaptive in context, for example avoiding or clinging to caregivers.

Third, early experience guides later behaviors and feelings via internal working models of attachment—"internal" because they reside in the mind, "working" because they guide perceptions and behaviors, and "models" because they are cognitive representations of relationship experiences. In other words, children store knowledge about relationships, especially knowledge about safety and danger, in models that guide their future interactions. Each new interpersonal interaction is processed and interpreted according to children's representations.

These models are assumed to operate, for the most part, outside of conscious experience. Knowledge gained from interactions with primary caregivers, typically parents, is of greatest importance; for example, children with loving parents develop positive models of relationships based on trust. Simultaneously, children develop parallel models of themselves; for example, children with loving parents view themselves as worthy of care. These models are assumed to generalize from parents to other people in children's lives, including friends and teachers. So, a child will assume that a friend or teacher is trustworthy if the child's primary caregiver is trustworthy.

Mary Dinsmore Salter Ainsworth and the Strange Situation

Ainsworth conducted the first observational studies of mothers and children that were rooted in attachment theory, first in Uganda and later in Baltimore, Maryland. Through her careful field notes, she noticed important individual differences among infants. Most appeared soothed by their mothers, while others were not, and still others displayed little emotion to their mothers' presence or absence. Ainsworth moved her work to the laboratory in order to assess the effect of maternal absence on infant exploratory behaviors. Her paradigm, called the Strange Situation, is a thirty-minute procedure that consists of a series of separations and reunions among a caregiver, a child, and a stranger.

Ainsworth and her students identified three patterns of attachment that were particularly evident from children's behavior in the reunion episodes with mothers. Most children displayed a pattern of attachment that Ainsworth and colleagues labeled "secure." When their mothers were present, these children displayed a balance between exploring the laboratory playroom and seeking proximity with their mothers. During separations, secure children displayed some distress as indicated, for example, by crying. When reunited, these children greeted their mothers warmly, often with hugs, and were easily soothed by them. Children classified as "insecure-ambivalent" displayed few exploratory behaviors when their mothers were present, often clinging to them. These children were usually very upset during separations. When reunited, they displayed angry and resistant or ambivalent behaviors toward their mothers. For example, they would cry and raise their arms to be picked up and then push their mothers away while continuing to cry. Children classified as "insecure-avoidant" explored the playroom when their mothers were present. Unlike other children, however, these children paid little attention to their mothers. In addition, these children were usually not upset during separations and snubbed or avoided their mothers during reunions. Mary Main and Judith Solomon identified a fourth pattern of attachment, "insecure-disorganized," characterized by extreme distress over separations and disorganized, disoriented, and confused behaviors during reunions. Specifically, these children displayed frozen postures, repetitive movements, and dazed facial expressions when reunited with their mothers.

Overwhelmingly, the Strange Situation has become the preferred method of assessing attachment in infancy. There is, in fact, considerable evidence that security status in the Strange Situation is related to parenting behaviors, especially maternal sensitivity, which can be defined as the mother's ability to perceive an infant's signals accurately and to respond promptly and appropriately. Children whose mothers are sensitive to their needs are likely to be classified as secure. Children with avoidant patterns tend to have mothers who are either rejecting or intrusive and overstimulating. Children with ambivalent patterns tend to have mothers who are inconsistent in their parenting behaviors; for example, they may be sensitive and responsive some of the time but not always, which makes it difficult for children to predict their behavior. Children with disorganized patterns tend to have mothers who have experienced loss, trauma, or mental illnesses.

Although most of the research that has been conducted on patterns of attachment concerns infants' relationships with their mothers, there is some work that has examined infants' relationships with their fathers. There is no debate that children develop full-fledged attachment relationships with their fathers. In other words, it is clear that children can and do develop multiple attachment relationships. Little is known, however, about how children integrate the knowledge gained from multiple attachment models, especially when the models are different. Yet, there is some evidence for concordance across attachment figures—children who are securely attached to their mothers are also likely to be securely attached to their fathers. Concordance is best explained by shared parenting values, although infant temperament has also been suggested as an explanation.

Child Care

By the twenty-first century, most infants in the United States experienced some form of child care in their first year of life. This represented an enormous shift in how children in the United States were raised, a shift that led to concerns about whether infant child care disrupts mother-child attachment. Some have argued that infants experience daily separations as maternal rejection, which should lead to avoidance, while others have suggested that separations prevent mothers from having sufficient opportunities to develop sensitive caregiving styles. The results of the National Institute of Child Health and Human Development Study of Early Child Care, a study of more than 1,000 infants and their mothers, clearly demonstrated that neither security nor avoidance in the Strange Situation was associated with type of care, amount of care, or quality of care. Instead, security was associated with characteristics of mothering, such as sensitivity. Infants who experienced dual risks, for example poor quality child care and insensitive mothering, were at increased risk for developing insecure attachments. Thus, the effects of child care on attachment depend primarily on the nature of ongoing interactions between mothers and children.

Other Measures of Attachment

The Strange Situation continues to be the benchmark method for assessing attachment security in infancy. Alternatives, however, have been developed. The Attachment Q-sort, developed by Everett Waters, is a method designed to assess attachment security naturalistically in the home environment. Observers sort a set of ninety cards with behavioral descriptions—for example, "Actively solicits comforting from adult when distressed"—from most characteristic to least characteristic of the child. The child's profile is compared to that of a prototypical securely attached child, based on attachment researchers' hypothetical sorts or rankings of the cards.

Methods have also been developed for assessing attachment security in adolescence and adulthood. Preeminent among these is the Adult Attachment Interview (developed by Carol George, Nancy Kaplan, and Mary Main), a semistructured interview in which adults are asked to reflect and report on their early experiences with attachment figures, typically their mothers and fathers. The coding system focuses on the consistency and coherency of responses. Adults are classified as "secure/autonomous" if they express value for their early attachment relationships and are able to report on these experiences in a clear and organized fashion. Adults are classified as "dismissing" if they devalue the importance of early attachment relationships by expressing disregard for negative experiences, by having few memories of childhood, or by having idealized memories of their childhoods. Adults are classified as "preoccupied" if they display confusion or anger regarding early attachment relationships and talk excessively about their early experiences concerning them. Finally, adults are classified as "unresolved/disorganized" if they demonstrate lapses in reasoning during discussions of loss or abuse.

Stability of Attachment and Later Relationship Functioning

There is some evidence that attachment status is a stable phenomenon, as evidenced by concordance between security in the Strange Situation during infancy and in the Adult Attachment Interview during adolescence or early adulthood. Specifically, secure infants become autonomous adults, while avoidant infants become dismissing and ambivalent infants become preoccupied. Instability in attachment classifications over time seems to be linked to salient life events. Events that may redirect secure infants toward patterns of insecurity in adolescence and adulthood include maltreatment, the loss of a parent, parental divorce, or a serious illness for the parent or child.

Strange Situation classifications in infancy are also predictive of later relationship functioning. Infants classified as secure show more positive emotions toward their parents at two years of age and have better communication with their parents during middle childhood than infants classified as insecure. Patterns of attachment in infancy are also predictive of the quality of relationships with people other than parents. For example, children who are securely attached to their caregivers have better relationships with teachers, peers, and close friends.

Clinical Implications

The field of attachment began with Bowlby's clinical work with disordered patients. Since then, researchers have remained interested in links between early attachment history and the development of psychopathology. Work with institutionalized children demonstrates that the failure to form attachment relationships can lead to serious mental health problems. Most research, however, concerns associations between the quality of care children receive from attachment figures and later behavior. For example, infants with ambivalent attachment relationships are more likely to develop later anxiety disorders, while those with disorganized attachment relationships are more likely to develop later dissociative disorders, where individuals lose touch with reality. There is little evidence for specific links between types of insecurity and types of disorders. Instead, insecurity seems to operate as a risk factor that is neither a necessary nor a sufficient cause for disorders. Not surprisingly, there is evidence that a secure attachment relationship functions as a protective factor for children; in other words, security may protect children from the effects of other risk factors associated with psychopathology, such as their own difficult temperaments.

The processes through which early attachment relationships lead to later disorders are not well understood. Most theorists, however, believe that internal working models must moderate any link between the two. Models characterized by anger, mistrust, anxiety, and fear may lead children not only to behave aggressively but also to interpret the behaviors of others, even kind behaviors, negatively. In fact, the early memories of people with personality disorders are characterized by marked distortions and inconsistencies that reflect their negative attributions of themselves and others. More research on internal working models, especially with respect to their resistance to change, could help direct future therapeutic efforts with both children and adults.

Cross-Cultural Research

Because attachment theory is grounded in evolutionary biology, one of its core assumptions is that infant-caregiver attachment is a universal phenomenon. This assumption is controversial. At the very least, however, research from around the world supports the claim that all infants develop attachment relationships, secure or insecure, with their primary caregivers. Beyond this, there is considerable evidence that the number of children who develop a secure pattern of attachment is proportionately similar across cultures. In African, Chinese, Israeli, Japanese, Western European, and American cultures alike, most children, about two-thirds, are securely attached to their caregivers.

The proportion of children who are insecure-avoidant or insecure-ambivalent, however, varies across cultures. Consider that in Japan a higher proportion of children are classified as ambivalent and a lower proportion of children are classified as avoidant than in Western European and American cultures. Japanese infants, in fact, are more likely to be very upset during separations from their caregivers and less likely to explore the environment than American infants. Based on these data and using the Japanese culture as an example, Fred Rothbaum and his colleagues offered a critique of the universality of attachment that focused on cultural variations in caregiver sensitivity and child competence.

Rothbaum and his colleagues argued that caregiver sensitivity in Japan is a function of parents' efforts to maintain high levels of emotional closeness with their children, but that in the United States it is a function of parents' efforts to balance emotional closeness with children's assumed need to become self-sufficient. In fact, Japanese parents spend more time in close contact with their infants than parents in the United States. Regardless, most attachment researchers now agree that caregiver sensitivity is only one important contributor to attachment security. In all cultures, other factors such as how much stimulation parents provide their children, as well as child characteristics such as temperament, are likely to influence the development of attachment.

The link between attachment security and child competence has also received scrutiny from a cross-cultural perspective. Child characteristics that are associated with security in Western cultures, such as independence, emotional openness, and sociability, are less valued in other cultures. Attachment security may lead to social behaviors that vary across cultures but are nonetheless adaptive in context. For example, Japanese secure children may be more likely than Western secure children to depend on others to meet personal needs, because interpersonal dependency is valued in the Japanese culture. In other words, the characteristics of child competence may differ across cultures as a result of culture-specific pressures.

See also: AINSWORTH, MARY DINSMORE SALTER; BOWLBY, JOHN; PARENT-CHILD RELATIONSHIPS

Bibliography

Ainsworth, Mary D. S., Mary C. Blehar, Everett Waters, and Sally Wall. Patterns of Attachment: A Psychological Study of the Strange Situation. Hillsdale, NJ: Erlbaum, 1978.

Bowlby, John. "The Influence of Early Environment in the Development of Neurosis and Neurotic Character."International Journal of Psychoanalysis 21 (1940):154-178.

Bowlby, John. Attachment and Loss, Vol. 1:Attachment. New York:Basic, 1969.

Bowlby, John. Attachment and Loss, Vol. 2:Separation. New York:Basic, 1973.

Bowlby, John. Attachment and Loss, Vol. 3:Loss. New York: Basic, 1980.

Bretherton, Inge. "Attachment Theory: Retrospect and Prospect."In Inge Bretherton and Everett Waters eds., Growing Points of Attachment Theory and Research. Chicago: University of Chicago Press, 1985.

Crick, Nicki R., and Kenneth A. Dodge. "A Review and Reformulation of Social Informational-Processing Mechanism in Children's Social Adjustment."Psychological Bulletin 115 (1994):74-101.

DeWolff, Marianne S., and Marinus H. van IJzendoorn. "Sensitivity and Attachment: A Meta-analysis on Parental Antecedents of Infant Attachment." Child Development 68 (1997):571-591.

Erikson, Erik. Childhood and Society. New York: Norton, 1950.

Fox, Nathan A., Nancy L. Kimmerly, and William D. Schafer. "Attachment to Mother/Attachment to Father: A Meta-analysis." Child Development 62 (1991):210-225.

Freud, Sigmund.An Outline of Psychoanalysis. New York: Norton, 1949.

George, Carol, Nancy Kaplan, and Mary Main. "The Adult Attachment Interview." Manuscript, University of California, Berkeley, 1996.

Greenberg, Mark T. "Attachment and Psychopathology in Childhood." In Jude Cassidy and Phillip Shaver eds., Handbook of Attachment: Theory, Research, and Clinical Applications. New York: Guilford Press, 1999.

Harlow, Harry F. "The Nature of Love." Amercian Psychologist 13(1958):573-685.

Main, Mary, and Judith Solomon. "Procedures of Identifying Infants as Disorganized/Disoriented during the Ainsworth Strange Situation." In Mark Greenberg, Dante Cicchetti, and E. Mark Cummings eds., Attachment in the Preschool Years: Theory, Research, and Intervention. Chicago: University of Chicago Press, 1990.

National Institute of Child Health and Human Development. Early Child Care Research Network. "The Effects of Infant Child Care on Infant-Mother Attachment Security: Results of the NICHD Study of Early Child Care."Child Development 68 (1997):860-879.

Rothbaum, Fred, John Weisz, Martha Pott, Kazuo Miyake, and Gilda Morelli. "Attachment and Culture: Security in the United States and Japan."American Psychologist 55 (2000):1093-1104.

Thompson, Ross A. "Early Attachment and Later Development."In Jude Cassidy and Phillip R. Shaver eds., Handbook of Attachment: Theory, Research, and Clinical Applications. New York: Guilford Press, 1999.

van Ijzendoorn, Marinus H., and Abraham Sagi. "Cross-Cultural Patterns of Attachment: Universal and Contextual Dimensions." In Jude Cassidy and Phillip R. Shaver eds., Handbook of Attachment: Theory, Research, and Clinical Applications. New York: Guilford Press, 1999.

Waters, Everett. "The Attachment Q-Set." In Everett Waters, BrianE. Vaughn, German Posada, and Kiyomi Kondo-Ikemura eds., Caregiving, Cultural, and Cognitive Perspectives on Secure-Base Behavior and Working Models. Chicago: University of Chicago Press, 1995.

Waters, Everett, Nancy S. Weinfield, and Claire E. Hamilton. "TheStability of Attachment Security from Infancy to Adolescence and Early Adulthood: General Discussion." Child Development 71 (2000):703-706.

Kathleen McCartney

Eric Dearing