Attachment refers to the special bond that develops between the infant and his or her primary caregiver and provides the infant with emotional security. Many psychologists believe that the quality of attachment has lifelong effects on our relationships with loved ones. Once attached, babies are distressed by separation from their caregiver (separation distress or anxiety). There is evidence that seven- to nine-month-old infants in every culture studied show distress when they are separated from their primary caregiver (Grossman & Grossman, 1990 ).
Bowlby’s Theory of Attachment
Bowlby’s ( 1969 ) evolutionary theory of attachment proposes that infants must have a preprogrammed, biological basis for becoming attached to their caregivers. This innate behavioral repertoire includes smiling and cooing to elicit physical attachment behaviors on the part of the caregiver. He argues that the attachment relationship between caregiver and child functioned as a survival strategy: Infants had a greater chance of survival if they remained close to the caregiver for comfort and protection.
Attachment as a survival strategy is illustrated in a study in Nigeria of Hausa infants and their caregivers (Marvin, VanDevender, Iwanaga, LeVine, & LeVine, 1977 ). The researchers report that the attachment relationship protected infants from the dangers of their environment, which included open fires, tools, and utensils that were easily accessible. Infants explored their environment, but only when they were in close proximity to an attachment figure. Similarly, among the Dogon of Mali, infants were always kept in close proximity with the mother (being held most of the time) and infants did not roam freely, thus avoiding dangers such as open fires, snakes, and animal droppings (True, Pisani, & Oumar, 2001 ).
Bowlby and Ainsworth’s Classification System of Attachment
Based on Bowlby’s attachment theory, Mary Ainsworth’s ( 1967 , 1977 ) study of mothers and infants in Uganda led to the tripartite classification system of attachment relationships. Based on her careful observations of 26 mother-infant pairs over a span of one year, she described three attachment styles: secure , ambivalent , and avoidant . The latter two attachment styles she labeled as “insecurely attached.” Infants who are securely attached become distressed when their mother leaves but are easily comforted by her when she returns. Infants who are ambivalent also experience distress when their mother leaves but when she returns they send mixed signals—they want to be comforted by her yet, at the same time, appear to have a difficult time letting her soothe them. Infants who are avoidant do not seem to be distressed when their mother leaves and when she returns these infants will actively avoid reuniting with their mother and instead focus their attention elsewhere. Ainsworth later replicated her results in a sample of Baltimore mothers and their infants. In her samples, she found that approximately 57 percent of mothers and infants were classified as securely attached, 25 percent as ambivalent, and 18 percent as avoidant.
Studies from other cultures have found a similar distribution of attachment classifications; others have found considerable differences. And some attachment styles are not reported in certain cultures. For example, no avoidant infants were found in True et al.’s ( 2001 ) study of the Dogon of Mali. Mali mothers kept their infants close to them throughout the day and practiced constant, responsive nursing (nursing on demand when the infant is hungry or distressed). This type of caregiving, True et al. argued, “prevents” avoidant attachment to the mother. These findings highlight the importance of understanding the attachment system in the context of parenting practices specific to each culture.
Cross-Cultural Studies on Attachment
Since Ainsworth’s early studies, hundreds of studies of attachment have been conducted in cultures all over the world. To measure attachment, the Strange Situation, developed by Ainsworth, is the most widely used measure. In the Strange Situation, infants are separated from their mothers for a brief period of time. The separation is thought to trigger the attachment system in the infant. The quality of attachment is derived partly from an assessment of the infant’s reaction to the separation and subsequent reunion with the mother.
Although this method has been used extensively across cultures, the cross-cultural validity of this method and the meaning of the attachment classifications themselves have been questioned. For instance, the meaning of the separation may differ across cultures (Takahashi, 1990 ). Japanese infants are rarely separated from their mothers, and the separation during the Strange Situation may represent a highly unusual situation that may imply something different for Japanese infants and their mothers than for U.S. infants and their mothers. It may also be the case that subtle attachment behaviors (for instance, those that characterize avoidant relationships) are difficult even for well-trained coders to observe in infants from different cultures (Crittenden, 2000 ; van IJzendoorn & Sagi, 1999 ).
In addition to criticisms leveled at the measurement of attachment, cross-cultural researchers have questioned the appropriateness of the different categories of attachment. For instance, researchers studying Chinese infants and their mothers question the validity of the avoidant category as an indication of insecure attachment (Hu & Meng, 1996 , cited in van IJzendoorn & Sagi, 1999 ). Chinese mothers emphasize early independence in their infants and, at the same time, stress their reliance on nonparental (usually the grandparent) caregivers. These factors, rather than an insecure relationship between the mother and her infant, may be responsible for why babies show avoidant attachment in China.
Van IJzendoorn and Sagi ( 1999 ) outline several important cross-cultural issues that Ainsworth’s Uganda study raised, including whether maternal sensitivity is a necessary antecedent of attachment. For instance, mothers of securely attached infants are described as sensitive, warm, and more positive in their emotional expression. Mothers of avoidant children are suspected of being intrusive and over-stimulating. Mothers of ambivalent children have been characterized as being insensitive and uninvolved. Thus, according to Ainsworth, a major determinant of attachment security is having a caregiver who is sensitive and responsive to the child’s needs. In a review of 65 studies of attachment, however, caregiver sensitivity was related only modestly to security of attachment (DeWolff & van IJzendoorn, 1997 ). And studies with other cultures found an even weaker connection between parent sensitivity and security of attachment (van IJzendoorn & Sagi, 1999 ).
One possible reason for why maternal sensitivity has not been consistently linked to secure attachment is that sensitivity may mean different things and expressed in different ways across cultures. One study contrasted U.S. caregivers’ with Japanese caregivers’ sensitive responsiveness (Rothbaum, Weisz, Pott, Miyake, & Morelli, 2000 ). In the United States, parents tend to wait for their child to communicate a need and then respond to that need. In other words, sensitive parenting in the United States allows the child to express his or her individual needs to the parent so that the parent can appropriately address those needs. In contrast, in Japan, parents tend to anticipate their child’s needs instead of waiting for their child to communicate a need. This can be done by being aware of situations that may cause distress to a child and anticipating ways to minimize the stress. Rothbaum and colleagues argue that researchers need to pay more attention to how different cultures conceptualize and demonstrate sensitive parenting to better understand what type of parenting leads to secure attachment. Further, parenting behaviors that, from a Western perspective, may seem to promote insecure attachment, may in fact do the opposite in other cultures. For example, Ainsworth suggested that parenting that is “intrusive,” namely, directive and controlling, leads to an insecure attachment. However, this type of parenting may have an entirely different meaning in non-Western cultures (Chao, 1996 ). Future research should consider the interaction between culturally specific parenting practices and infant temperament and behaviors to better understand how attachment relationships develop (Keller, 2008 ).
Is Secure Attachment a Universal Ideal?
In the United States, secure attachment is assumed to be the ideal. The very term that Ainsworth and colleagues chose to describe this type of attachment, and the negative terms used to describe others, reflects this underlying bias. Some research suggests that cultures may differ, however, in their notion of “ideal” attachment. For example, northern German mothers value and promote early independence and see the “securely” attached child as “spoiled” (Grossmann, Grossmann, Spangler, Suess, & Unzner, 1985 ). Children raised in traditional Japanese families are also characterized by a high rate of ambivalent attachment, with practically no avoidant types (Miyake, Chen, & Campos, 1985 ). Traditional Japanese mothers seldom leave their children (such as with babysitters) and foster a strong sense of dependence in their children. This dependence supports the traditional cultural ideal of family loyalty. As noted earlier, being separated, even for a short time (such as during the Strange Situation) may be highly unusual and thus it is more likely that children’s behavior in this situation be recorded as insecurely attached. Attachment scholars have suggested that researchers should stop using value-laden terms such as “secure” and “insecure” in describing the attachment relationship. Instead, they propose that it may be more useful to describe the attachment relationship as “adaptive” or “maladaptive” to the specific context, which would take into consideration how cultures differ in the particular attachment strategy that may be most appropriate for that culture (Crittenden, 2000 ; Keller, 2008 ).
Although cross-cultural studies of attachment have been criticized, there is still a keen interest in attachment relationships. One main reason for this interest is that attachment relationships in childhood may have long-term consequences—into adolescence and adulthood. A compilation of the most important longitudinal studies of attachment from Germany, Israel, England, and the United States, suggests that early attachment relationships are related to the quality of later peer relationships, the ability to develop close, intimate adult relationships, and how one parents (Grossman, Grossman, & Waters, 2005 ). These studies also show, however, that there is a complex relation between early attachment relationships and later development that depends on the temperament and personality of the child, the stability and characteristics of the caregiving environment, individual life events, and quality of interpersonal relationships throughout the child’s life.
In sum, the vast literature concerning attachment in different cultures suggests that attachment between infants and their caregivers is a universal phenomenon. What may differ across cultures, however, is the specific attachment behaviors exhibited by the infant that indicate secure or insecure attachment and the parenting practices that promote secure and insecure attachment (van IJzendoorn & Sagi, 1999 ). Further, as stated earlier, attachment scholars argue that viewing attachment through the lens of being “adaptive” and “maladaptive” may be more useful than using the evaluative terms “secure” and “insecure” (Crittenden, 2000 ; Keller, 2008 ). “Adaptive” attachments, then, would refer to relationships that promote the maximum level of safety for the child within a specific cultural context. Researchers could then define an “optimal” relationship between infant and caregiver as one that may be achieved in different ways, under different circumstances, in different cultures.
TEMPERAMENT AND ATTACHMENT: A SUMMARY
Much still needs to be done to understand the attachment patterns in other cultures and the relationship among cultural milieu, infant temperament, and attachment style. Notions about the quality of attachment and the processes by which it occurs are qualitative judgments made from the perspective of each culture. What is considered an optimal style of attachment in one culture may not necessarily be optimal across all cultures. Furthermore, because nonparental caretaking is a frequent form in most cultures (Weisner & Gallimore, 1977 ), examining the attachment “network” instead of focusing solely on dyads, as has traditionally been done, is of crucial importance (van IJzendoorn & Sagi, 1999 ). Studies involving an African tribe of forest-dwelling foragers known as the Efe, for instance, show a very different pattern from the one psychologists have come to accept as necessary to healthy attachment (Tronick, Morelli, & Ivey, 1992 ). Efe infants are cared for by a variety of people in addition to their mothers; the time spent with caregivers other than their mothers increases from 39 percent at 3 weeks to 60 percent by 18 weeks. Efe infants are always within earshot and sight of about 10 people and they have close emotional ties to many people other than their mothers. However, when infants are one year old, they clearly show a preference for being cared for by their mothers and become upset when left by their mothers. At this age, then, mothers once again become the primary caretakers. Thus, there is evidence that attachment to a primary caregiver is still formed and that children are emotionally healthy despite having multiple caregivers. Nonetheless, despite the fact that in many cultures there are multiple caregivers other than the infant’s biological mother, studies of attachment that focus beyond the mother, are still relatively rare.