This study was funded by a grant from the Spencer Foundation to E. Bates and I. Bretherton. We would like to thank the 30 mothers for their cooperativeness and patience in responding to our many probing questions. We also thank the children for supplying the data. In addition we wish to express our special gratitude to Carol Williamson and Sandra McNew for help in collecting and analyzing the data.Requests for reprints should be sent to Inge Bretherton,

No abstract

Research based on maternal observations shows that many children acquire their first feeling-state words at around 18-20 months (Bretherton, McNew, & Beeghly-Smith, 1981;Bretherton et al., 1986). Mothers also report that by 28 months, most 132 This document is copyrighted by the American Psychological Association or one of its allied publishers.This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Attachment theory is based on the joint work of John Bowlby (1907Bowlby ( -1991 and Mary Salter Ainsworth (1913-). Its developmental history begins in the 1930s, with Bowlby's growing interest in the link between maternal loss or deprivation and later personality development and with Ainsworth's interest in security theory. Although Bowlby's and Ainsworth's collaboration began in 1950, it entered its most creative phase much later, after Bowlby had formulated an initial blueprint of attachment theory, drawing on ethology, control systems theory, and psychoanalytic thinking, and after Ainsworth had visited Uganda, where she conducted the first empirical study of infantmother attachment patterns. This article summarizes Bowlby's and Ainsworth's separate and joint contributions to attachment theory but also touches on other theorists and researchers whose work influenced them or was influenced by them. The article then highlights some of the major new fronts along which attachment theory is currently advancing. The article ends with some speculations on the future potential of the theory.

Bowlby postulated that transactional patterns between caregiver and infant become internalized by the infant as “internal working models” of self and other in relationship and that these working models then determine how the infant interprets the caregiver's behavior and responds to it. When parent and infant or child are not reciprocally responsive to signals, defensive processes may interfere with the adequate development and functioning of working models in the child. Not only does this affect the observed relationship, it also influences the way in which an individual (adult or child) discusses attachment relationships with a third person. Corroboration for this view comes from work with adults (the Adult Attachment Interview, the Parent Attachment Interview) and children (the Separation Anxiety Test, the Attachment Story Completion Procedure). If, as research suggests, insecure parents' working models of attachment relations are distorted by defensive processes, the resulting insensitive behaviors toward the child may interfere with the child's construction of adequate working models, thus providing a potential explanation for the intergenerational transmission of insecure attachment relations in those cases where the parent has not been able to work through a rejecting or neglecting attachment relationship experienced in childhood.