Bowlby's ‘trilogy’ - Attachment, Separation, and Loss - had, by 2010, been cited over 12 000 times. By this measure at least, John Bowlby is the most influential psychoanalyst of all time, establishing a new developmental paradigm. Yet Bowlby's initial aims were relatively modest. He wanted to observe and understand the responses of children separated from their mothers and their developmental sequelae. Attachment lays out, in typical Bowlbian fashion, a logical, exhaustive account demolishing existing theories of the parent-infant bond - psychoanalytic and behavioural - clearing the way for his own novel model. Drawing on neo-Darwinistic ethology and control systems theory, he proposed attachment as a primary motivational force. The attachment dynamic is a product of our ‘environment of evolutionary adaptation’ in which protection from predation was essential. The ‘set goal’ of attachment is physical proximity to a secure base when a child is threatened, stressed, or ill. Only once attachment needs are assuaged, can exploration and play resume.

Attachment theory's co-creator was Mary Ainsworth. Separation is largely devoted to the implications of the ‘strange situation’, which she devised, in which individual differences in 1-year-olds' responses to brief separations from parents are reliably observed and classified. The familiar concepts of secure and insecure attachment and its subtypes, insecure-avoidant, insecure-ambivalent and insecure-disorganised, flow from this work. Ainsworth established that maternal sensitivity and responsiveness in the first year of life was the key to later security. Longitudinal studies have shown how social competence is linked to early security of attachment, supporting the psychoanalytic hypothesis of the determining influence of early relationships.

In Separation Bowlby analysed anxiety disorders, especially agoraphobia, in terms of insecure attachment. He saw anger as a normal and adaptive response to separation, a negative reinforcement schedule aimed at maintaining proximity. The clinical manifestations of anger, including self-harm, and the rage of people with borderline personality disorder, can be seen as pathological attempts to restore a modicum of security in the face of overwhelming threat. Loss was strongly influenced by two attachment-influenced sets of research paradigms: Parkes' studies of bereavement, and Brown and Harris's studies of loss and depression. The presence of a secure base and the quality of attachment crucially determine whether depression or recovery is the outcome of loss or death.

Bowlby's trilogy remains the secure foundation for half a century of ‘normal science’ (the post-paradigm phase). Mary Main identified disorganised attachment as a risk factor for severe psychopathology, devising the Adult Attachment Interview as a systematic measure of mental representations of attachment. Fonagy and colleagues have extended her model into the concept of ‘mentalising’ and its deficiency in psychopathology.

Some paradigm shifts feel so familiar that one thinks ‘why had no one thought of that before?’ The trilogy stirred up much opposition, but attachment is now part of the air we breathe. Bowlby bemoaned the divide between ‘biological’ and psychodynamic psychiatry, insisting that his ethological-developmental model was rooted in evolutionary biology. Today's epigenetics and neuropsychoanalysis suggest that divide is narrowing. Psychiatry is on the brink of another paradigm shift; one that Bowlby would have welcomed.