Attachment theory is among the most influential ideas in developmental psychology and among the most misrepresented in parenting content.
The core observation
Infants form a bond with their primary caregivers and use them as a base from which to explore and to return for comfort.
Which was established through observation and through a structured procedure examining how infants respond to brief separation and reunion.
Patterns of response were categorised, and those categories became the framework.
What the categories describe
Secure attachment, where a child seeks comfort on reunion and is comforted.
Insecure patterns, categorised by whether the child avoids the caregiver or is difficult to settle.
And a further category involving contradictory behaviour, associated with more concerning circumstances.
Which describe a relationship rather than a child, and the same child can show different patterns with different caregivers.
What predicts security
Sensitivity — noticing a child's signals and responding appropriately — is the factor most consistently identified.
Which is not the same as constant availability or immediate response, and it is not the same as never letting a child be distressed.
Meta-analyses find the association between sensitivity and attachment security is real and moderate rather than overwhelming, which means many other things also matter.
What is overstated
The popular version frequently implies that specific practices — how a baby is carried, fed, or settled — determine attachment.
Which the research does not support. The findings concern the overall pattern of responsiveness rather than particular practices.
Attachment parenting, as a named approach, is a separate set of practice recommendations that borrowed the term rather than following from the research.
Stability
Attachment classification in infancy predicts later outcomes with modest effect sizes.
Which is a genuine association and is far from determinative, and classifications change over time in a substantial proportion of children as circumstances change.
Which means early experience matters and does not fix outcomes.
Multiple caregivers
Children form attachments with several caregivers, and these can differ in quality.
Which contradicts the assumption of a single critical relationship, and it means fathers, grandparents and consistent childcare providers are attachment figures in their own right.
Research on childcare has found that quality of care and stability of caregivers matter, and that childcare use itself does not disrupt attachment to parents.
Cross-cultural questions
The structured procedure was developed in one cultural context, and its interpretation across cultures has been questioned substantially.
Which is a live debate — behaviour that reads as insecure in one setting may reflect normal child-rearing practice in another.
Researchers in the field have written directly about the misapplication of the framework in policy and legal settings.
The practical version
Responding to a child, most of the time, with reasonable accuracy, is what the research supports.
Which allows for getting it wrong frequently, since repair after rupture appears to matter as much as avoiding rupture.
Anyone concerned about their relationship with their child, or about their own capacity to respond, should speak to a health visitor or doctor, since effective support exists.
Rupture and repair
Research examining interaction sequences finds that mismatches between caregiver and infant are extremely frequent, and that repair follows most of them.
Which reframes the goal — not constant attunement, which does not occur, but a pattern where things go wrong and are put right.
The proportion of interactions that are coordinated at any moment is far lower than parents assume, in observational studies of typical dyads.
Applications in policy
Attachment concepts have been used in family court and child protection decisions, and researchers in the field have written directly about misapplication.
Which includes use of the structured procedure outside its validated context and interpretation of categories as fixed characteristics of children.
Position statements from research groups have set out what the evidence does and does not support for these purposes.
Fathers and partners
Research on father-child attachment has grown and finds similar processes with some differences in the interactions associated with security.
Which means the framework is not specific to mothers, despite the original research focus.
Play and challenging interaction appear more prominent in the paternal literature, though the distinction may reflect typical roles rather than anything inherent.
Sleep and attachment
Claims that specific sleep arrangements determine attachment are made in both directions and are not supported by the research.
Which means neither bed sharing nor sleep training has been shown to determine attachment security.
Studies examining behavioural sleep interventions have found no differences in attachment measures at follow-up, which is directly relevant to a frequently made claim.
Adoption and fostering
Children who experienced early adversity can form secure attachments with new caregivers.
Which is a consistent finding, and the process may take longer and require specific support.
Therapeutic approaches for this exist, and some approaches marketed under attachment terminology have been criticised as harmful and are not endorsed by professional bodies.
Anyone seeking support in this area should check what evidence and professional endorsement an approach has.