We were interested to read Dr O'Connell's response to our paper. Unfortunately we did not have space to do justice to a discussion of reactive attachment disorder from an evolutionary point of view, although we agree that this is an important theoretical perspective. Dr Minnis first became interested in reactive attachment disorder when working as an orphanage doctor in Guatemala. Most of the children there displayed symptoms of the disinhibited form of the disorder and it seemed clear that these behaviours were adaptive in a setting where primary attachment figures were lacking. We have touched on the maintenance of these behaviours from an evolutionary perspective in a previous paper (Reference Minnis, Marwick and ArthurMinnis et al, 2006).

Dr O'Connell also points out that we did not engage in a discussion of attachment theory, or the work of John Bowlby (Reference BowlbyBowlby, 1973). We do not wish to underestimate the crucial role of Bowlby's work in advancing our understanding of childhood development, however, we were unable to do justice to the complex interplay between attachment patterns and reactive attachment disorder within the space allowed. This important topic is the focus of our prevous publication (Reference Minnis, Marwick and ArthurMinnis et al, 2006). In short, children can be securely attached while suffering from reactive attachment disorder and children suffering from the disorder have difficulties in various domains of early development, not simply the domain of attachment (Reference Richters and VolkmarRichters & Volkmar, 1994; Reference Green and GoldwynGreen & Goldwyn, 2002). Research into reactive attachment disorder is in its infancy and is a field ripe for exploration on a number of fronts.