Insecure Attachment 267respondents resemble children with type A attachments – attachment
behaviours and feelings are denied, restricted or repressed.
2 Autonomous. The respondent values attachment relationships and either
gives a convincing history of emotionally supportive relationships in
childhood or has come to terms with a childhood lacking them (called
by some ‘earned security’). Typically self-reliant, objective, and non-
defensive, these respondents resemble children with type B attach-
ments – attachment feelings and behaviours are expressed in open and
balanced ways, with appropriate dependency and confidence in the
attachment figure.
3 Preoccupied. The account of childhood relationships is confused, incoher-
ent and unobjective. They are still caught up in childhood events and
unable to move beyond them. Anger at parents is unresolved. These
respondents resemble children with type C attachments – attachment
feelings and behaviours are exaggerated and ambivalent.
4 Unresolved-disorganised. Discussions of potentially traumatic events are
marked by striking lapses in the monitoring of reasoning or discourse,
suggesting dissociated memory systems or abnormal absorption in unre-
solved traumatic memories. These respondents resemble children with
type D attachments – disorganised.
Given the complexity and seeming subjectivity of this rating scheme,
it is truly remarkable how well parent and infant attachment status
correspond. When the parent’s attachment type is classified using the
AAI (whether before or after the child’s birth), and when the infant’s
attachment to a parent is classified as A, B, C or D using the SSP, then
roughly two-thirds of infants match their parents in attachment type. This
concordance is all the more striking given the difference in assessment
methods between analyses of transcripts of what adults say and analyses
of videotapes of how infants behave.
Consequences of secure and insecure attachments
Many studies have compared the social and psychological development of
securely and insecurely attached children. The picture that has consistently
emerged is that securely attached children do better on average. However,
not all securely attached children do well, and not all insecurely attached
children do badly. It is still unclear whether this group difference arises
because insecure attachment is itself the key risk factor; the alternative
is that insecure attachment is simply a marker for wider-ranging family
abnormalities that have the adverse long-term effects. Recent research
suggests that children can do well despite being insecurely attached,
so long as other things are favourable (good parenting overall, stable
education, no unpleasant traumas, etc). Vice versa, children with secure
attachment patterns who are brought up in poverty and fail to make good
friendships, etc. typically do much less well. Thus, a lesson is that while
attachment patterns are important, they are not wholly deterministic and