Dear Sir,

Peter Fonagy's Guest Editorial (‘Memory and therapeutic action’, Int. J. Psychoanal., 80:215-23) compellingly links the psychoanalytic tradition, current attachment literature, and cognitive ideas on the psychology of memory. From an attachment perspective it can be extended in a number of ways.

First, while the distinction between declarative and implicit memory is crucial, it is also important to consider what might be the relationship between these two memory systems. Attachment research suggests that secure and insecure developmental pathways tend to be relatively stable, so that children diagnosed as insecure in infancy will be similarly classified at school entry and subsequently. The mechanisms maintaining this stability are only just beginning to be discovered, but it presumably depends on consistent parent-child dyads in which each evokes a particular set of secure-or insecure-making responses from the other.