Attachment and the internal world of children at risk: the South African context

Abstract

Attachment research and clinical practice has tended to focus on four traditional attachment types (i.e. secure, avoidant, ambivalent and disorganized). However, for thinking and practice to advance it is critically important that these categories and their assessment be interrogated in different contexts and across developmental age. Furthermore, similarities between attachment and psychoanalytic concepts, particularly between internal working models and objects relations, are being articulated in the psychoanalytic literature. A group of theorists have positioned themselves as psychoanalytic attachment theorists, bridging attachment theory and psychoanalytic thinking. However, there has been misunderstanding on both sides which needs to be clarified (Fonagy, 1999; 2018). Thus, this empirical research explored possible links between the internal world and attachment development in pre-adolescent children. To answer these questions, a quantitative design was used with descriptive and inferential statistics, and general linear models were employed to analyze the data. Questionnaires were administered to 105 children between the ages of eight and twelve from socio-economically deprived contexts with high exposure to trauma. These questionnaires were the Attachment Story Completion Test, Thematic Apperception Test, Differential Emotions Scale-IV, Social Cognition and Object Relations Scale G and the Senior South African Individual Scale – Revised, Similarities subtest. A staggering 93% of the children had insecure attachments (specifically, avoidant 37%, disorganized 34% and ambivalent 22%) and attachment was more complex than currently conceptualized in the attachment literature. Primary attachment did not only present as one of four discreet categories, but 70% of the sample had a complex attachment with more than one type of attachment style present. Attachment Intensity also varied along a continuum (between 1 and 5). Relationships between attachment type and developmental markers (namely: object relations, intensity of emotion and defense styles) were largely not significant in this sample. However, Attachment Intensity was found to: a) moderate the relationship between secure and ambivalent attachments and Positive Emotionality; and b) between attachment

iv complexity and Emotional Investment in Relationships; and c) influence Complexity of Representations of People, and Identity and Coherence of Self. Generally secure attachments were better aligned with object relations than insecure attachments, although disorganized attachments influenced Identity and Coherence of Self. Thus, while points of overlap between attachment and object relations has been foregrounded in the literature, these results contest reflexive theoretical interpretations (Fonagy & Campbell, 2015) that suggest identical overlap between these two constructs, and cast new light on how the interaction between attachment and object relations, intensity of emotion and defense styles can be understood. The findings have implications for research that has not considered complexity of attachment or attachment intensity. Findings also have policy implications for supporting mother-infant dyads, particularly in high risk contexts.