‘What babies need’ was the focus for the observations which gave rise to the science of attachment. Babies need a secure one-to-one bond with their mums. Should they not form such a bond, they face a darkened future.

John Bowlby was the man who made this vision famous. His book Attachment (1969) is all about the biology of what babies need and why they need it, what social behaviours they are born with, and how these come to focus on a single mother-figure over their first three years of life. How mothers do or should behave made up a very small part of Bowlby’s book. It was all about babies.

Bowlby imagined that prehuman babies had only survived by ‘attaching themselves’ to their mums. Evolution has made sure that newborns now all have a set of highly stereotyped and observable ‘attachment behaviours’ which are ‘emitted’ when the baby feels vulnerable – behaviours like smiling, crying, looking, clinging, babbling and sucking. By the end of year one, all these behaviours will become focused on one adult, said Bowlby. From here the child would go on to form a thermostat-like ‘attachment behavioural system’ which, in times of trouble, made ‘proximity to mum’ its set-goal.

Over time,these claims have all proven false, as documented in our book Babies in Groups. In fact, young babies’ social behaviours only ‘promote proximity’ to adults if adults care about babies and read their behaviours as signs of need. So long as babies grow up surrounded by caring attentive adults, they will likely survive.

Despite his inaccurate claims, Bowlby’s focus on infant behaviour should have moved science towards the overdue recognition that babies’ capacities and needs merit investigation in their own right. But his initiative has fallen flat. 55 years on and babies’ social exploits have, once again, all but vanished from attachment-inspired research. Even the ‘strange situation procedure’ – which is widely used to classify toddlers’ attachments – avoids coding babies’ observable behaviours, purporting to divine instead various categories of ‘inner organization’ of ‘attachment security.’ The labels which result are primarily used to judge how mothers mother, not to illuminate how babies live and act. Those lively bundles of emotion and intelligence, who delight in sights and sounds and smells, and in playing and working with other babies, have all but fallen from view.

So, we arrive at a paradox. The attachment paradigm, which began as a push to highlight babies, their capacities and their needs, has now gone into reverse, transferring its attention to the rights and wrongs of maternal behaviour. The baby’s side of attachment has today withered to unanchored speculations about children’s predisposition to develop ‘internal working models’ of their relationships with adults. These hypothetical cognitive constructs are, at best, invisible, ill-defined, and, even in theory, only achieved long after babies have grown into walking talking children.