Attachment Theory

Attachment Theory Definition

An attachment refers to the strong emotional bond that exists between an infant and his or her caretaker. The attachment theory is designed to explain the evolution of that bond, its development, and its implications for human experience and relationships across the life course. Although attachment theory has primarily been a theory of child development, since the 1980s, the theory has had a large impact on social psychological theories of close relationships, emotion regulation, and personality.

Attachment Theory History and Background

Attachment TheoryAttachment theory was originally developed by John Bowlby, a British psychoanalyst who was attempting to understand why infants experience intense distress when separated from their parents. Bowlby noticed that separated infants would go to extraordinary lengths (e.g., crying, clinging, frantically searching) to either prevent separation from their parents or to reestablish contact with a missing parent. At the time, psychoanalytic writers held that these behaviors were expressions of immature defense mechanisms that were operating to repress emotional pain, but Bowlby observed that such expressions may serve an evolutionary function.

Drawing on evolutionary theory, Bowlby postulated that behaviors such as crying and searching were adaptive responses to separation from a primary attachment figure—someone who provides support, protection, and care. Because human infants, like other mammalian infants, cannot feed or protect themselves, they are dependent upon the care and protection of older and stronger adults. Bowlby argued that, over the course of human evolution, infants who were able to maintain proximity to an attachment figure (i.e., by looking cute or by expressing in attachment behaviors) would be more likely to survive to a reproductive age. According to Bowlby, a motivational-control system, what he called the attachment behavioral system, was gradually crafted by natural selection to help the child regulate physical proximity to an attachment figure.

The attachment behavior system is an important concept in attachment theory because it provides the conceptual link between evolutionary models of human development and modern theories on emotion regulation and personality. According to Bowlby, the attachment system essentially “asks” the following question: Is the attachment figure nearby, accessible, and attentive? If the child perceives the answer to this question to be “yes,” he or she feels loved, secure, and confident, and, behaviorally, is likely to explore his or her environment, play with others, and be sociable. If, however, the child perceives the answer to this question to be “no,” the child experiences anxiety and, behaviorally, is likely to exhibit attachment behaviors ranging from a simple visual search for the parent to active following and vocal signaling. These behaviors continue until either the child is able to reestablish a desirable level of physical or psychological proximity to the attachment figure, or until the child “wears down,” as may happen in the context of a prolonged separation or loss. In such cases, Bowlby believed that the child may develop symptoms of depression.

Individual Differences in Infant Attachment Patterns

Although Bowlby believed that his theory captured the way attachment operates in most children, he recognized that there are individual differences. Some children, for example, may be more likely to view their parents as inaccessible or distant, perhaps because their parents have not been consistently available in the past. However, it was not until Bowlby’s colleague, Mary Ainsworth, began to study infant-parent separations that individual differences in attachment were incorporated formally into attachment theory. Ainsworth and her students developed a technique called the “strange situation”—a laboratory paradigm for studying infant-parent attachment. In the strange situation, 12-month-old infants and their parents are brought to the laboratory and are separated from one another and then reunited. In the strange situation, most children (i.e., about 60%) behave in the way implied by Bowlby’s “normative” understanding of attachment. Specifically, they become upset when the parent leaves the room, but when he or she returns, they actively seek the parent and are easily comforted by him or her. Children who exhibit this pattern of behavior are often called “secure.” Other children (about 20% or less) are ill-at-ease initially and, upon separation, become extremely distressed. Importantly, when reunited with their parents, these children have a difficult time being soothed and often exhibit conflicting behaviors that suggest they want to be comforted but that they also want to “punish” the parent for leaving. These children are often called “anxious-resistant.” The third pattern of attachment that Ainsworth and her colleagues documented is called “avoidant.” Avoidant children (about 20%) don’t appear too distressed by the separation and, upon reunion, actively avoid seeking contact with their parent, sometimes turning their attention to play objects on the laboratory floor.

Ainsworth’s work was important for at least three reasons. First, she provided one of the first empirical demonstrations of how attachment behavior is patterned in both safe and frightening contexts. Second, she provided the first empirical taxonomy of individual differences in infant attachment patterns. According to her research, at least three types of children exist: those who are secure in their relationship with their parents, those who are anxious-resistant, and those who are anxious-avoidant. Finally, she demonstrated that these individual differences were correlated with infant-parent interactions in the home during the first year of life. Children who appear secure in the strange situation, for example, tend to have parents who are responsive to their needs. Children who appear insecure in the strange situation (i.e., anxious-resistant or anxious-avoidant) often have parents who are insensitive to their needs, or inconsistent or rejecting in the care they provide.

Attachment in Adult Romantic Relationships

Although Bowlby was primarily focused on understanding the nature of the infant-caregiver relationship, he believed that attachment played a role in human experience across the life course. It was not until the mid-1980s, however, that psychologists began to take seriously the possibility that attachment may play a role in adulthood. Cindy Hazan and Phil Shaver were two of the first researchers to explore Bowlby’s ideas in the context of romantic relationships. According to Hazan and Shaver, the emotional bond that develops between adult romantic partners is partly a function of the same motivational system— the attachment behavioral system—that gives rise to the emotional bond between infants and their care-givers. Hazan and Shaver noted that infants and care-givers and adult romantic partners share the following features:

  • Both feel safe when the other is nearby and responsive.
  • Both engage in close, intimate, bodily contact.
  • Both feel insecure when the other is inaccessible.
  • Both share discoveries with one another.
  • Both play with one another’s facial features and exhibit a mutual fascination and preoccupation with one another.
  • Both engage in “baby talk” or “motherese” (i.e., a high-pitched, idiosyncratic language that involves “made up” words that only the couple understands).

On the basis of these parallels, Hazan and Shaver argued that adult romantic relationships, like infant-caregiver relationships, are attachments, and that romantic love is a property of the attachment behavioral system, as well as the motivational systems that give rise to caregiving and sexuality.

Three Implications of Adult Attachment Theory

The idea that romantic relationships may be attachment relationships has had a profound influence on modern research on close relationships. There are at least three critical implications of this idea. First, if adult romantic relationships are attachment relationships, then we should observe the same kinds of individual differences in adult relationships that Ainsworth observed in infant-caregiver relationships. Second, if adult romantic relationships are attachment relationships, then the way adult relationships “work” should be similar to the way infant-caregiver relationships work. Third, whether an adult is secure or insecure in his or her adult relationships may be a partial reflection of his or her attachment experiences in early childhood.

Do we Observe the Same Kinds of Attachment Patterns Among Adults That We Observe Among Children?

If adult romantic relationships are attachment relationships, then the same kinds of individual differences that Ainsworth observed in infant-caregiver relationships should be observed in adult relationships. Some adults, for example, may be expected to be secure in their close relationships—feeling confident that their partners will be there for them when needed and being open to depending on others and having others depend on them. Other adults, in contrast, will be expected to be insecure in their relationships. For example, some insecure adults may be anxious-resistant: They worry that others may not love them completely, and they may be easily frustrated or angered when their attachment needs go unmet. Others may be avoidant: They may appear not to care too much about close relationships and may prefer not to be too dependent upon other people or to have others be too dependent upon them.

The earliest research on adult attachment involved studying the association between individual differences in adult attachment and the way people think about their relationships and their memories for what their relationships with their parents are like. In 1987, Hazan and Shaver developed a simple questionnaire to measure these individual differences. (These individual differences are often referred to as “attachment styles,” “attachment patterns,” “attachment orientations,” or “differences in the organization of the attachment system.”) In short, Hazan and Shaver asked research subjects to read three paragraphs and indicate which paragraph best characterized the way they think, feel, and behave in close relationships. Paragraph A described discomfort and nervousness in being close to others, as well as difficulty with trust and intimacy; paragraph B depicted relative ease with closeness to and mutual dependence with others; and paragraph C indicated a perception that others are hesitant to get close as desired and that a partner doesn’t love them or likely won’t want to stay with them.

Hazan and Shaver found that the number of people endorsing each of these descriptions was similar to the number of children classified as secure, anxious, or avoidant in infancy. In other words, about 60% of adults classified themselves as secure (paragraph B), about 20% described themselves as avoidant (paragraph A), and about 20% described themselves as anxious-resistant (paragraph C).

Do Adult Romantic Relationships “work” in the Same way That infant-Caregiver Relationships work?

If adult romantic relationships are attachment relationships, then the way adult relationships “work” should be similar to the way infant-caregiver relationships work. For the most part, research suggests that adult romantic relationships function in ways that are similar to infant-caregiver relationships. Naturalistic research on adults separating from their partners at an airport demonstrated that behaviors indicative of attachment, such as crying and clinging, were evident and that the way people expressed those behaviors was related to their attachment style. For example, while separating couples generally showed more attachment behavior than nonseparating couples, people with avoidant attachment styles showed much less attachment behavior.

There is also research that suggests that the same kinds of features that mothers desire in their babies are also desired by adults seeking a romantic partner. Studies conducted in numerous cultures suggest that the secure pattern of attachment in infancy is universally considered the most desirable pattern by mothers. Adults seeking long-term relationships identify responsive caregiving qualities, such as attentiveness, warmth, and sensitivity, as most attractive in potential dating partners. Despite the attractiveness of secure qualities, however, not all adults are paired with secure partners. Some evidence suggests that people end up in relationships with partners who confirm their existing beliefs about attachment relationships, even if those beliefs are negative.

Are Attachment Patterns Stable From infancy to Adulthood?

An important implication of attachment theory is that whether an adult is secure or insecure in his or her adult relationships may be a partial reflection of his or her attachment experiences in early childhood. Bowlby believed that the mental representations or working models (i.e., expectations, beliefs, “rules,” or “scripts” for behaving and thinking) that a child holds regarding relationships are a function of his or her caregiving experiences. For example, a secure child tends to believe that others will be there for him or her because previous experiences have led him or her to this conclusion. Once a child has developed such expectations, he or she will tend to seek out relational experiences that are consistent with those expectations and perceive others in a way that is colored by those beliefs. According to Bowlby, this kind of process should promote continuity in attachment patterns over the life course, although it is possible that a person’s attachment pattern will change if his or her relational experiences are inconsistent with his or her expectations. In short, if we assume that adult relationships are attachment relationships, it is possible that children who are secure as children will grow up to be secure in their romantic relationships.

Research shows that there is a modest degree of overlap between how secure people feel with their mothers, for example, and how secure they feel with their romantic partners. For example, among people who are securely attached to their mothers, over 65% of them are likely to feel secure with their romantic partners too. There is also evidence suggesting that people who are secure as children are more likely to grow up to become secure adults. Of secure adults, approximately 70% of them were classified as secure when they were 12 months of age in the strange situation. Taken together, these data suggest that there is a moderate degree of stability in attachment styles from infancy to adulthood, but that there is also plenty of room for people’s ongoing experiences to shape their security.


  • Hazan, C., & Shaver, P. R. (1994). Attachment as an organizational framework for research on close relationships. Psychological Inquiry, 5, 1-22.