Waters, Merrick et al., (2000) highlight that attachment theory predicts both stability under ordinary circumstances and change when negative life events alter caregiver behaviour, and defines these negative life events as “loss of a parent, parental divorce, life-threatening illness of parent or child, parental psychiatric disorder, and physical or sexual abuse by a family member”. This study found that the infant-mother attachment relationship was significantly related to attachment security twenty years later and also found that most often changes from secure to insecure attachment were associated with stressful life events. However this was not always the case and the relationship between life events and attachment patterns across time was not perfect. This finding does suggest a strong level of support for the validity of attachment theory over time however it raises questions about why some people change classifications and others remain stable regardless of whether or not they are affected by negative life events.
A study by Bar-Haim, Sutton, Fox & Marvin, 2000 found that both stability and change of attachment over time are significantly affected by negative life events. In this study the quality of attachment at 14, 24 and 58 months was found to be similar to other studies of children at comparable ages (e.g. Ainsworth et al., 1978; Belsky, Rovine & Taylor, 1984; Stevenson-Hinde & Shouldice, 1993; Turner, 1991; van Ijzendoorn & Kroonenberg, 1988 – as cited in Bar-Haim et al., 2000) however this study divided participants into two groups. One group was referred to as having “stable-secure” attachment across time and the other group included those who experienced a “change” in attachment classification. Mothers of children in the “change” group reported significantly higher rates of negative life events compared to mothers of children in the “stable-secure” group and conversely mothers of children in the “stable-secure” group reported more positive life events in comparison to mothers in the “change” group. These findings suggest that both change and stability are more likely to be related to the current state of the parent-child relationship rather than the attachment quality during infancy and that continuity in classification is probably due to stability in maternal behaviour and family environment.
Much of the evidence supporting attachment theory has been conducted on middle-class samples offering both advantages and disadvantages. Middle class participants are representative of a large proportion of the population, however as Waters, Merrick et al. (2000) acknowledge stability in middle-class samples may be reflective of more than inherent stability of attachment security. Low rates of negative experiences related to attachment and strong social support structures assisting these participants to cope with negative experiences may contribute to the stability of secure attachment in this sample. Equally, high levels of stressful events in disadvantaged samples may contribute to the stability of insecure attachment for those participants.
Lamb (1987) shows the clearest evidence for “the long-term predictive validity of Strange Situation classifications comes from the ambitious Minneapolis studies”. Lamb highlights one major problem however affecting generalisability and interpretation of the findings. Stability in family circumstances was consistent among participants and was either assumed or actively sought through selection procedures. The results may therefore suggest that experiences during infancy do not necessarily have enduring effects on the child because the data may imply that it is actually continuity in parenting which is more likely to influence development because early patterns of behaviour are more likely to be maintained. If this is the case, it suggests that recent rather than early patterns of parent-child interaction may determine the differences in attachment stability and that the long-term effects of early attachment are less influential than predicted by attachment theory.
Thompson (2000) found that children vary considerably in the extent to which early attachment relationships have an enduring impact, highlighting that reasons for individual patterns of continuity or change are not clear. This study found that changing family circumstances and how they impact on the quality of parenting influence the extent to which early attachments are enduring. “Broader family processes and the adult’s adaptive capacities may either exacerbate or buffer the impact of family stresses on early child-parent attachment” (Thompson, 2000). To further interpret these findings exploring the child’s temperamental vulnerability or adaptability may shed some light in understanding how family events impact the attachment relationship.
Temperamental theories highlighting the role of anxiety levels
While attachment theorists have made important contributions to current views of early experience and individual differences, temperament theorists argue that more attention should be paid to individual temperamental properties as these play a vital role in determining a child’s vulnerability to anxiety levels, which in turn affects how their quality of attachment is viewed in the Strange Situation.
Kagan (1984) argues that infants with high and low levels of vulnerability to anxiousness will be viewed as insecurely attached. Infants with high levels of anxiousness would become extremely distressed by the Strange Situation and infants with low levels of anxiousness would not be frightened by the unfamiliar (or the Strange Situation), with both groups consequently being classified as insecure. However infants who are only moderately vulnerable to anxiety are most likely to be classified as securely attached because they are the most easily placated.
Advances in neuroscience have stimulated a revival of interest in temperament and have allowed scientists to gain an understanding of some forms of behavioural variation. Kagan & Sidman (1991) assert that the disposition for inhibited or uninhibited behaviour is a temperamental characteristic which may be under partial genetic control, and that infants more vulnerable to anxiety are inhibited and those less vulnerable to anxiety are uninhibited. This study found that infants who are likely to become inhibited “have a lower threshold of excitability in the circuits that involve the amygdala and the sympathetic nervous system” (Kagan & Sidman, 1991) than do those who are likely to be uninhibited. These findings are consistent with evidence from studies conducted on older inhibited and uninhibited children.
Attachment and temperament theorists do have some common ground and both recognise that children differ from birth and that parent-child relationships are influenced by both parent and child characteristics and outside influences. The attachment theorist’s position however is quite deterministic and does not place enough emphasis on the nature of the child and puts too much emphasis on the nurturing environment. The temperament theorists view seems a better fit with the current view of both nature and nurture interacting inseparably. Kagan & Sidman (1991) assert that a person’s psychological profile is neither permanently fixed by biology nor shaped entirely by social interaction. Kagan (1984) acknowledges that although differences in rearing environments do make substantial contributions to individual differences, an infant’s particular temperamental style profoundly influences their treatment by others and their reactions to unexpected situations. Temperament theorists do not argue that mother-child relationships are irrelevant to development, they do however see them as less important, but at the same time concede that care giving behaviours during infancy do influence development.
Cultural appropriateness of the Strange Situation in assessing attachment security
Research on attachment theory has primarily been conducted in Western countries who share comparable cultural values and who view positive outcomes similarly. However different cultures value different outcomes and questions on the validity of patterns of attachment cross-culturally are raised.
Patterns of attachment may be remarkably different in countries with different culture attitudes and patterns of child-care. Studies conducted on infants in North Germany, Japan and those in Kibbutz care in Israel found that attachment theory could not be reliably applied cross culturally (Goldberg, 2000). These studies found a very different distribution of attachment patterns compared to those found in middle class North American samples. This is not to say that the concept of attachment is not important because attachment relationships occur and impact upon humans universally, however measuring attachment using the Strange Situation during infancy in these cultures raises interesting questions.
Goldberg (2000) highlights the Grossman et al., (1985) study in which forty nine percent of North German children were classified as insecure-avoidant using the Strange Situation. This contrasts to American babies where twenty six percent were classified in this category in Ainsworth’s Baltimore study. Does this data actually infer that many more North German children are insecurely attached? A more likely explanation may be that mothers conform to cultural expectations in their style of infant care because Germans value more highly early independence, and actively seek to teach their children not to rely on too much assistance from others.
Evidence cited by Goldberg (2000) from a study of Kibbutz-reared infants in Israel also showed significantly more insecurely attached infants than was found in Ainsworth’s Baltimore sample, while studies on Japanese infants from Sapporo revealed no avoidant infants but an increase in resistant attachments. It must be remembered however that Japanese infants are infrequently separated from their mother so the Strange Situation is more likely to be excessively stressful for these infants.
The above studies show limitations of the Strange Situation requiring consideration in weighing up how valid the assessment is in regard to cultural influences and practices.
Other considerations impacting on developmental issues and associated later outcomes
Recognition must be given to the fact that much research does support the validity of attachment theory, however the framework needs broadening. While empirical findings show attachment theory to usefully measure and predict later outcomes on the dimension of security, most research measures only the dimension of attachment security and does not look at other aspects of the parent-child relationship (Dunn, 1993). As Dunn (1993) points out, we need to measure different aspects of children’s relationships with their parents to shed further light on how different features of the relationships might be related to differing later outcomes.
Relationship aspects not measured but possibly important for the impact upon attachment quality include emotional and cognitive aspects considered important in adult relationships. There are also many developmental changes in children which impact on parent-child relationships and as a child’s development dramatically improves for understanding and relating to their world their relationships also change in significant ways (Dunn, 1993). Attachment researchers have paid little attention to “possible developmental changes in individual differences in quality of attachment” (Crittenden, 1990 – as cited in Dunn, 1993), “with the exception of the striking evidence that life events that affect the family can alter children’s status” (Vaughn, Egeland, Sroufe & Waters, 1979; Lamb, Thompson, Gardner & Charnov, 1985 – as cited in Dunn, 1993). Looking at other dimensions to see how developmental changes affect the pattern of individual differences in parent-child relationships is important to further understanding stability and change in attachment over time.
Conclusion
It is clear that developmental psychology must not underestimate the enormous contribution that attachment theory has made to understanding socioemotional behaviours or to the importance of the attachment bond during infancy. Recognising that secure attachment relationships are a huge contributor enabling people to explore the world during all stages of life, but also appreciating the adaptability and flexibility of the human brain to adjust to different circumstances, it is not obvious from the evidence reviewed here that being securely attached as an infant is beneficial to later socioemotional functioning indefinitely. The evidence reviewed for this essay found mixed results in relation to stability of attachment, with some studies indicating relative stability over time, and other studies showing significant instability. The evidence suggests that a secure attachment during infancy is more likely to produce positive later outcomes, however consistently the research questions whether later outcomes are actually the result of a secure infant attachment relationship, or whether later outcomes are related to other variables such as temperament and stable family and environmental circumstances. Evidence here reveals that early experiences are most predictive of later socioemotional outcomes only when the combination of very stable environmental conditions and family stability are present. Overall, the evidence seems to be more heavily weighted to suggesting that there is more change in attachment over time than attachment theory suggests, and that while the attachment bond during infancy in important, it is not realistic to view it as ‘crucial’ for later outcomes.
Bibliography
Bar-Haim, Y., Sutton, D.B., Fox, N.A. & Marvin, R.S. (2000). Stability and change of attachment at 14, 24, and 58 months of age: Behavior, representation, and life events. Journal of Child Psychology & Psychiatry & Allied Disciplines, 41(3), 381-388.
Berlin, L.J., Cassidy, J. & Belsky, J. (1995). Loneliness in young children and infant-mother attachment: A longitudinal study. Merrill-Palmer Quarterly, 41(1), 91-103.
Calkins, S.D. & Fox, N.A. (1992). The relations among infant temperament, security of attachment, and behavioural inhibition at twenty-four months. Child Development, 63(6), 1456-1472.
Dunn, J. (1993). Young children’s close relationships: Beyond attachment. California: Sage.
Kagan, J. (1984). The nature of the child. NewYork: Basic Books.
Kagan, J. (1992). Yesterday’s premises, tomorrow’s promises. Developmental Psychology, 28(6), 990-997.
Goldberg, S. (2000). Attachment and Development. New York: Arnold.
Lamb, M.E. (1987). Predictive implications of individual differences in attachment. Journal of Consulting and Clinical Psychology, 55(6), 817-824.
Parkes, C.M., Stevenson-Hinde, J. & Marris, P. (1991). Attachment across the life cycle. London: Routledge.
Rutter, M. (1995). Clinical implications of attachment concepts: Retrospect and prospect. Journal of Child Psychology & Psychiatry & Allied Disciplines, 36(4), 549-571.
Santrock, J.W. (2001). Child Development (9th ed.). New York: McGraw-Hill.
Thompson, R.A. (2000). The legacy of early attachments. Child Development, 71(1), 145-152.
Waters, E., Hamilton, C.E. & Weinfield, N.S. (2000). The stability of attachment security from infancy to adolescence and early adulthood: General introduction. Child Development, 71(3), 678-683.
Waters, E., Merrick, S., Treboux, D., Crowell, J. & Albersheim, A. (2000). Attachment security in infancy and early adulthood: A twenty-year longitudinal study. Child Development, 71(3), 684-689.
Is the attachment bond during infancy crucial to later outcomes?
Michelle Quee