1.12.2 Bowlby attachment theory

Bowlby's evolutionary account

Attachment is a lasting bond with a caregiver who gives comfort in distress and a secure base for exploration.

• Bowlby held that the human infant is born helpless and needs a caregiver to survive, so attachment behaviour is an evolved, adaptive system for protection.

• Attachment is driven by an innate need for proximity and security, not by feeding: Bowlby rejected the secondary-drive or "cupboard love" account.

• Distress, pain, illness, fatigue, unfamiliarity and separation switch the system on, and comfort settles it.

• A frightened child stops exploring, and a child who feels secure explores.

• Children stay attached to caregivers who maltreat them, so attachment does not show that care is safe.

Evidence from animals

• Harlow's surrogate mother experiments in infant rhesus monkeys showed that contact comfort, not feeding, is the basis of the bond: infants clung to the soft cloth surrogate even when only the wire surrogate gave milk.

• Imprinting, described by Lorenz in goslings and drawn on by Bowlby, is rapid learning in a critical period after hatching, in which the young animal comes to follow the first moving object it meets.

Tap to enlarge

Monotropy and multiple attachments

• The principal attachment figure is defined by consistent, responsive care, not by kinship or sex, so it may be a father, grandparent or other carer rather than the biological mother.

• Monotropy is Bowlby's term for the infant's innate bias towards one principal attachment figure, ranked above the subsidiary attachments to other carers and different in kind from them.

• Polytropy is attachment spread across a network of several caregivers, each of whom the child misses when they leave and each of whom can soothe the child.

• By about 18 months most infants are attached to more than one familiar figure: in Schaffer and Emerson's study about 87% had multiple attachments and about 13% only one.

• Attachment describes a relationship, so a child can be secure with one caregiver and insecure with another.

Phases of attachment development

Bowlby's four phases are told apart by whether the child discriminates familiar people, protests at separation and takes account of the caregiver's plans.

• The boundary between the first two phases is also given as 8 to 12 weeks, or 2 months.

• A specific attachment to one figure, shown by stranger wariness and separation protest, first appears at about 7 to 9 months; clear-cut attachment is clearest after 7 months.

• Attachment behaviours peak at about 6 to 18 months.

• The sensitive period for forming a first selective attachment runs from about 6 to 24 months, matching the clear-cut phase.

• These ages are guides, not diagnostic cut-offs.

• The social smile first appears at about 6 to 8 weeks, after the earlier reflex smile and before laughing aloud at 3 to 4 months; its absence by 8 weeks is a red flag needing assessment.

Stranger and separation anxiety

Stranger anxiety starts once the infant can tell familiar from unfamiliar faces.

Internal working models

• The internal working model is Bowlby's construct: paired representations of whether others will be available and whether the self is worthy of care, built from real caregiving experience.

• Expected comfort encourages direct signals of distress, expected rejection encourages suppression, and inconsistent responses encourage stronger signals.

• These models often work outside awareness, and later relationships, including therapy, can change them.

• Attachment behaviour persists across the lifespan into adult and therapeutic relationships, less intense after early childhood but never extinguished, carried forward by internal working models.

• Ainsworth's sensitivity, noticing a child's signal, understanding it and responding promptly and appropriately, is moderately associated with security; video feedback that improves it can improve security.

The Strange Situation

• Mary Ainsworth devised the Strange Situation, the standard laboratory assessment of infant attachment; Bowlby built the theory it tests.

• It was designed and validated for infants of about 12 to 18 months, when separation protest, stranger wariness and exploration are reliably present.

• Its eight short episodes include two separations and two reunions.

• Classification rests mainly on behaviour at the reunion episodes, especially the second, not at the separations.

Tap to enlarge

• Secure attachment is the majority in low-risk Western samples, at roughly 60 to 70%, about two in three infants.

• Avoidant and resistant patterns are organised strategies, so an avoidant infant still has a bond; disorganisation is a breakdown of strategy.

• Avoidant infants who look calm are physiologically aroused on heart rate measures.

• Main and Solomon added disorganised attachment to Ainsworth's original three, giving four infant categories: secure, avoidant, resistant and disorganised.

• Disorganisation is associated with maltreatment but does not prove that maltreatment occurred.

• Culture changes Strange Situation behaviour and the proportions classed avoidant or resistant.

What a classification predicts

• Insecurity is not a diagnosis: most insecure children develop no disorder, and links with later problems are modest and non-specific.

• Insecure attachment in early childhood is a risk factor for school-age behavioural problems, most often oppositional defiant disorder.

• Insecure-avoidant attachment in infancy is linked to later perpetrating bullying, through negative working models of others, poor emotion regulation and hostile social information processing.

• Disorganisation is associated with externalising problems and dissociation.

Adult Attachment Interview

The Adult Attachment Interview (AAI), by George, Kaplan and Main, classifies an adult's current state of mind about attachment from the coherence, content and language of their account of childhood, not from security with any one person. A mother's AAI category corresponds with her infant's Strange Situation classification.

• A speaker coded unresolved is also given the best-fitting organised category.

• Earned security is a harsh childhood coded autonomous because the account is coherent and reflective, often after supportive later relationships.

• Continuity of attachment from infancy into adulthood is moderate.

Separation

Robertson and Bowlby described a young child's response to prolonged separation from the attachment figure in three phases.

• Anaclitic depression is Spitz's description of depressive withdrawal with physical decline in an attached infant over about 6 months old who suddenly loses the attachment figure; it reverses on reunion.

• Familiar, responsive substitute care, prior security and a good reunion lessen the impact; repeated changes of caregiver do most harm.

Deprivation and privation

• Deprivation between about 6 months and 3 years, when selective attachments form, carries lasting risk that later good care may not reverse, whereas the same deprivation ending earlier does not.

• Bowlby's maternal deprivation hypothesis treated all separation from the mother as one experience and overstated its lasting harm.

Rutter separated the two experiences:

Institutional care and adoption

• The Bucharest Early Intervention Project is the only randomised trial of family foster care against continued institutional care for abandoned children; developmental gains were larger the earlier the placement, and benefit was real but partial.

• Hodges and Tizard followed children raised in London residential nurseries: the ex-institutional syndrome of problematic peer relatedness persisted at age 16, even when adoption produced good family attachments.

• In children adopted after early adversity, security recovers after placement, but pre-adoptive adversity leaves a lasting mark on attachment representations.

Attachment disorders

Both disorders follow grossly inadequate early care, need a developmental age of at least 9 months, and are distinct from insecure attachment.

• In the International Classification of Diseases, 11th revision (ICD-11), reactive attachment disorder cannot be diagnosed alongside autism spectrum disorder.

• ICD-11 does not allow reactive attachment disorder before 12 months of age (developmental age under 9 months), and 5 years is the upper limit for onset.

• Disinhibited social engagement can persist after a selective attachment forms.

Attachment and bonding

• The two can be discordant: strong parental love does not guarantee secure attachment, and delayed feelings do not predict insecurity.

• There is no established critical bonding window just after birth, and a difficult birth or neonatal admission does not mean the bond is damaged.

• Persistently absent feelings towards the baby warrant assessment for postnatal depression.