Attachment Theory and Psychoanalysis Peter Fonagy New York:  Other Press, 2001.  261 pp., £22.50

When I began my pre-clinical training in the early 80s the notion of attachment and psychoanalysis did not seem to be on the horizon. As Peter Fonagy says in his preamble to this important and comprehensive book, ‘There is bad blood between psychoanalysis and attachment theory’. Even today there is a tendency in some psychoanalytic circles to assert that Attachment Theory is at best a blunt instrument with coarse measurements compared with the fine distinctions and formulations possible in the psychoanalytic consulting room. There has often been an anxiety that by considering the potential complementarity of a distinct and different discourse there might be a takeover or a kind of dilution of psychoanalytic understandings rather than an open dialogue. It must be said that in other professional groups working with deprived children and adults there has been more interest in the concepts of the attachment paradigm than in those of psychoanalysis, but this should not cause psychoanalytic workers to pre-empt dialogue. Attachment Theory has valuable resources to offer which can be used alongside other thinking in the consulting room. Peter Fonagy points no fingers but simply indicates that there has been defensive thinking on both sides of the divide, with John Bowlby himself at times frustrating ‘a mutually corrective relationship’.

In a conference report on ‘What does empirical research contribute to psychoanalysis: The Example of Infant Research’ (Rustin, 1998) the author indicates that there was a heated, or perhaps icy, division, between infant researchers and some classical analysts, who repudiated the idea that observation of real children has any relevance to the child of the internal world: the child or the dream seemed to be the apparently irreconcilable dialectic here. While the conference explicitly cited infant observation research in the psychoanalytic tradition, the same resistance can still quite frequently be seen to attachment formulations. Fonagy uses t