On: Attachment and Psychoanalysis; Working Models and Part-Object Transferences

Dear Editors,

Wachtel applies attachment theory to psychoanalytic work, presenting a case and re-conceptualizing concepts of one- and two-person constructs (Wachtel, 2010). In our 1997 paper, The transference refracted through the lens of attachment (Szajnberg and Crittenden, 1997), we presented three cases to demonstrate the impact of working models of attachment on the transference over the course of psychoanalyses and the implications for interpretive technique. Each case presented with different working models of attachment early in the work, that is, internal representations of how the analysand expected intimate relations to operate during periods of anxiety: the first with Insecure, Enmeshed/Angry, the second with Insecure, Avoidant/Defended and the third with Secure attachment. But for the second case, an adolescent, we demonstrated how his working model of attachment shifted over time of the treatment from initial Insecure Avoidant to a raucous period of Insecure, Enmeshed/Angry to resolution with Secure attachment at the end of four years' treatment. (A more complete account is in, Szajnberg, 1993.) In more classical analytic terms, a first approximation is to describe these as shifts from paternal to maternal transferences in the early and midphases.

We suggested in 1997 that individuals may have ‘nesting’ models of attachment, building on Bowlby's idea of attachment constructs may vary with different parents, work that has been demonstrated empirically by subsequent researchers (Fonagy, 2001). This is a different way of conceptualizing and integrating attachment theory with psychoanalysis, specifically with object relations theory. Internal models of attachment may be different with representations of a father versus mother, and in a more complicated manner, with part-object representations of significant others. In the transference, therefore, this would be manifested during moments of anxiety, as part-object transferences towards the analyst, with their