A Psychiatrist Who Left Psychoanalysis
John Bowlby (1907–1990) was positioned to become a British psychoanalytic orthodox heir. But treating delinquent youth at London's Tavistock Clinic in the 1940s, he concluded Freudian drive theory — symptoms as frustrated libido — could not explain children's separation experiences.
He pulled in three outside disciplines. First, ethology of Konrad Lorenz and Niko Tinbergen — a gosling's 'imprinting' on the first moving object after hatching was not learning but an evolved behavioral system. Second, control systems theory — behavior modeled as homeostatic loops reducing distance to a 'set-goal.' Third, cognitive psychology's representations — children encode caregiver experiences as 'internal working models' carried for life.
This synthesis became Volume 1 of Attachment and Loss (1969). The central claim is uncompromising: attachment is not a byproduct of feeding but an evolved primary motivational system steering proximity-seeking under threat. Harry Harlow's 1958 rhesus infants — choosing cloth mother over wire-mesh feeder — provided lateral evidence.
Volume 2 Separation: Anxiety and Anger (1973) interpreted the protest-despair-detachment sequence ethologically; Volume 3 Loss: Sadness and Depression (1980) argued adult grief is the same system at work.
Bowlby's 1951 WHO report Maternal Care and Mental Health documented developmental damage in institutionalized infants, sparking controversy. Politicized into 'working mothers ruin children,' Bowlby's actual claim was narrower: consistent primary caregiver (not necessarily biological mother) bonds are essential.
Mary Ainsworth — Theory Into the Lab
Bowlby's theory was intuitively appealing but weakly measurable. Filling that gap was Canadian-born developmental psychologist Mary Ainsworth (1913–1999).
Following her husband to Uganda in 1954, she observed 28 infant-mother pairs at home for 9 months, publishing Infancy in Uganda (1967) — the first systematic data linking infant security to maternal responsiveness.
Moving to Johns Hopkins, she conducted 18 home visits of 4 hours each over a year for 26 Baltimore families and simultaneously designed a lab protocol. Her 1978 Patterns of Attachment with Blehar, Waters, and Wall canonized the Strange Situation Procedure (SSP).
The SSP places 12–18 month infants in an 8-episode (~20 min) lab sequence: enter with mother → stranger enters → mother briefly leaves → mother returns → leaves again → stranger returns → mother returns. The diagnostic moment is reunion behavior — not how much the infant cries at separation, but how she receives comfort when mother returns determines classification.
Four Patterns
| Classification | Separation | Reunion | Caregiver | US prevalence |
|---|---|---|---|---|
| Secure (B) | Some distress, less exploration | Active contact, quickly soothed, returns to play | Consistently sensitive, responsive | ~62% |
| Insecure-avoidant (A) | Minimal distress, apparent indifference | Avoids approach, looks away, retreats to toys | Rejecting, suppresses emotional expression | ~15% |
| Insecure-resistant (C) | Intense distress, exploration halts | Simultaneously seeks and pushes away, not soothed | Inconsistent, unpredictable responsiveness | ~9% |
| Disorganized (D, added 1986) | Contradictory, freezing, dissociative | No coherent strategy (approach-then-stop, aimless spinning) | Frightened/frightening, trauma history | ~14% |
D was added by Mary Main and Judith Solomon in Affective Development in Infancy (1986) for cases uncodable in Ainsworth's original taxonomy. It correlates most strongly with abuse, neglect, or unresolved caregiver trauma.
How Culture Varies — van IJzendoorn & Kroonenberg 1988
Is the American norm (B ~62%) universal? Leiden's Marinus van IJzendoorn and Pieter Kroonenberg meta-analyzed 32 samples across 8 countries (N≈2,000) in 1988.
Result: Secure was the majority everywhere, but insecure distributions varied. North German samples showed more avoidant (A ~35%); Israeli kibbutzim and Japanese samples showed more resistant (C). Interpretations diverged: Germany's early autonomy emphasis, Japan's near-constant maternal proximity making SSP separation acutely jarring. Korean infant studies (Jin et al. 2012) show similar broad patterns despite distributional differences.
Another meta-analytic finding: within-country variance exceeded between-country variance by 1.5x — SES and caregiving environment matter as much as 'national culture.'
Does Early Attachment Decide Everything? — Sroufe & Egeland
Alan Sroufe and Byron Egeland's Minnesota Longitudinal Study (1975–) tracked ~200 children of low-income mothers from birth to adulthood — a developmental science gem. 12-month SSP classification statistically predicted peer relationships, academic achievement, adolescent depression/conduct problems, and adult romantic relationships.
But effect sizes hardly amount to 'destiny.' Predictive r from 12-month classification to adult outcomes typically ranges 0.20–0.30. Not 'insecure attachment ruins your life,' but 'one variable interacting with other risk and protective factors.' Early security can collapse under later trauma; early insecurity can transform into 'earned secure' through later stable relationships (partner, therapist).
Critiques — Honestly
Kagan's temperament critique. Jerome Kagan's 1984 The Nature of the Child argued much of what SSP measures may be infant temperament, not caregiver-child relationship. High behavioral-inhibition infants react more to separation; low ones react less. Studies failing to separate caregiving from temperament may overestimate effects. Modern attachment research controls for temperament.
Rothbaum's cultural critique. Fred Rothbaum, John Weisz and colleagues argued in American Psychologist (2000) that 'secure = autonomous exploration from a safe base' reflects Western individualist values. East Asian cultures may value proximity and interdependence (Japanese amae) as healthy attachment, so SSP-classified 'resistant' Japanese infants might be culturally adaptive. The critique landed as a caution against using Western norms as universal benchmarks.
Cautious D interpretation. A 2017 consensus paper led by Pehr Granqvist warned against D classification being misused in clinical and legal contexts as 'evidence of abuse.' D is a risk signal, not a diagnosis; standalone use is inappropriate.
Korean Attachment Research
Korean research began in earnest in the late 1980s. Kang Ki-sook's 1989 dissertation applied SSP to Korean infants; Lee Young and others in the 1990s–2000s linked caregiving responsiveness to attachment. After 2000, as childcare centers spread with dual-earner families, the central question became how center-based care interacts with attachment. Findings are consistent: quality of responsiveness matters more than quantity of caregiving time. Full-day center-attending children show no significant secure-attachment gap from home-reared peers if parents are consistently responsive.
Korean adoption research is rich. Bae Ki-jo (2014) and others tracked domestically and internationally adopted Korean children, suggesting high 'earned security' likelihood when adopted before 6 months. In Korean developmental psychiatry, SSP serves as adjunct assessment in suspected abuse/neglect cases and infant depression evaluation.
Meaning: Scientific Foundations of Child-Rearing
What Bowlby and Ainsworth built is not a prescription for 'good parenting.' What they built was a measurable science of the hypothesis that human intimate relationships are an evolved behavioral system. On this foundation Hazan and Shaver's 1987 adult attachment extension (#259), 2000s neurobiology (oxytocin, CRF), and clinical models like EFT and MBT were constructed.
Before asking 'what is my attachment style?', know what the theory proved and what it did not. Secure attachment is not a parenting trophy — it is evolution's default expectation. And secure does not mean perfect: Donald Winnicott's 'good enough mother' suffices.
