We live in an age preoccupied by ‘risk’ and its assessment and management, and by its partner, ‘security’ - both global and personal. How refreshing to review a book which addresses some of the psychological roots, in terms of attachment theory, of our sense of internal and mental securities (and insecurities) and that of our patients and institutions, and specifically one consequence of failed security - social and interpersonal violence.

This volume, which is divided into theoretical, clinical, institutional and research sections, gathers together a body of original work on attachment applied to forensic psychiatry and psychotherapy, along with some previously published work.

Attachment theory originated in, and has since been developed from, the work of John Bowlby. It offers the possibility of (limited) quantification of mental representations that hitherto remained only qualitative, and can provide a bridge between the understanding provided by cognitive science and psychoanalysis.

The first (and longest) chapter, by Peter Fonagy, sets out densely but clearly the theoretical basis of developmental failure of emotional containment, of mothering and mirroring. The consequences include disorders of attachment, which in this view lead to relative failures of the capacity for mental reflection, affect discernment and regulation, to confusions of the psychic and somatic, and to later possible interpersonal, social violence. Psychiatrists may be more familiar with the resultant diagnostic categories than with the details of intrapsychic developmental processes; the common denominator is ‘borderline personality disorder’, which has been Fonagy's focus of study. Two further theoretical chapters address largely technical matters within attachment research.

Preliminary research data presented by Fonagy (on adult and young offenders), by Adshead & Bluglass (on factitious illness by proxy), by Ross & Pfäfflin fflin (on male prisoners and control groups, including, intriguingly, fundamentalist Christians) and by Lamott et al (on female killers and victims of domestic violence) all point to relative disorders of attachment - especially of the unstable and dismissive types. Lamott et al propose a new category, ‘fragmented attachment representation’ (FRAG), for the intensely disorganised narratives of some individuals - presumably as a result of their complex traumatic pasts.

The chapters on attachment representations and institutions - secure hospitals (Adshead as a psychotherapist, and Aiyebusi from a nursing perspective) and prisons (Parker & Morris) - should be of interest to all of us who work in these toxic environments. They offer helpful ways of conceptualising individual and organisational problems, and of creative responses within institutions which are, after all, crucibles of pathology as well as society's attempt at providing ‘security’ and, hopefully, repair.

However, treatment per se is only thinly represented in this volume. I would not count this as an omission if attachment studies were seen as derivative from - as merely a part of an adjunct to - more complete psychodynamic theories, but there seems to be some ambivalence or lack of clarity about this running through the book. Attachment theory is elevated in some places to another ‘analytic theory’, and in the main clinical contribution to the book - the chapter by Paul Renn - this is made manifest. This chapter is very well informed and clearly competent from an attachment theory perspective, but seemed to me to raise the perennially important issue of how to integrate exciting theoretical and methodological advances into clinical work without skewing the essential, open-minded and personal nature of the therapeutic relationship. This application of attachment theory focuses on, and seeks out, childhood ‘trauma’ - separation, loss and abuse - from the first meeting and, predictably, finds it. It is heavily agenda-driven and there must be a danger that it may be experienced as intrusive and coercive, not to say bleak - even if tolerated by compliance. It reminded me of those ‘recovered memory’ therapists of the 1990s, who were said to offer their clients the prospect of new memories, and invariably delivered. I am not sure what to think of the fact that this intervention was offered as ‘short-term counselling in a probation setting’, and was therefore obligatory.

The significance of the findings of attachment research, and their potential for persuasion in the political and policy arena, seem to me immense. Fonagy, at the end of his chapter, articulates well the personal and societal consequences of our changing social structures, and failure to support families and parenting function sufficiently for adequate childhood emotional development. In the terms of this book:

Meanwhile, those with or without personality disorders increasingly seek the solace of counselling; antisocial behaviour (and antisocial behaviour orders) are on the increase; crime (and the ‘fight against crime’) and, for example, domestic violence, intensify; and prisons are the most predictable of (rapid) growth industries. Is it too much to hope that a book such as this might cause a few policy makers to pause? I highly recommend this volume for clinicians, managers and policy makers alike.