I wonder if revisiting this discussion in the following frame might be at all useful. Dr. Kagan’s point as I understand it is that social and circumstantial factors are much better predictors of mental health than is early attachment history. Dr. Siegel’s point is that attachment history is useful and therapeutically illuminating in the clinical setting.

The frame that occurs to me is that the differing points of view of Drs. Kagan and Siegel mean that, even though they’ve reached a very rewarding and civilised agreement, they are talking about radically different tasks and mental frames. Dr. Kagan is taking an essentially epidemiological stance when he asks “what factors best allow me to predict the mental health of the next person to walk through my door?” Dr. Siegel is taking a clinical or, more properly, a therapeutic stance when he asks “how will I best make sense of the history of the next person to walk through my door in order to assist them in overcoming their particular mental/emotional difficulty?”

Obviously Dr. Kagan’s stance is seeking to apply knowledge across a population while Dr. Siegel’s stance is seeking an intellectual structure to give shape to therapeutic work with particular people with whom he works.

These two stances have radically different criteria of utility or validity. For Dr. Kagan, a construct must increase predictive accuracy by an appreciable amount over rival constructs; for Dr. Siegel the construct must assist him to work [more] effectively with his patients. Dr. Kagan’s criteria are quite objective. From the research on therapeutic factors that has been published over the last few years (Wampold, Bergin et al.), we know that the most valuable function of all therapeutic or psychological models is the confidence that they provide the practitioner using them and their plausibility to the client. We do more effective therapy when we have confidence in the model and interventions that we are using and the client does better when they find our confidence infectious.

With this slightly embarrassing finding in mind, in the therapeutic setting the factual accuracy of attachment theory as identifying the causes of particular dysfunction is largely irrelevant. Of course we do know that attachment history contributes to differences between people’s neurological development which is a major factor in its ability to inspire our confidence. However we know that Transactional Analysis can also be very effective and the “reality” of Parent Adult Child is not in any sense objective.

Another aspect of this interplay is that the population voluntarily seeking therapy is [probably] more homogeneous as to social factors than is the population at large, which seems to be Dr. Kagan’s area of focus. In this smaller population where the differential impact of social factors is comparatively low, other factors like specific trajectories through interpersonal attachment may have a greater weight in accounting for our clients’ self-evaluated mental comfort.

So the bottom line for me as a therapist from this discussion at the moment is a strong warning that attachment history is unlikely to be anything like a complete pointer to individual clients but it is a useful model in which to characterise their current and historical relations. Simon Mundy, Psychotherapist near Sydney, Australia.