Attachment

Introduction

Attachment is the deep, enduring emotional bond that forms between a child and their primary caregiver, in most cases within the first year of life. This topic examines how attachments form, the different types of attachment, evolutionary explanations, the effects of deprivation and institutionalisation, and how early attachment experiences influence later relationships.

Key Concepts

Animal Studies

Lorenz (1935) — Imprinting in geese:

Procedure: Lorenz divided a clutch of greylag goose eggs into two groups. One group hatched under normal conditions with the mother goose; the other hatched in an incubator with Lorenz as the first moving object they saw.

Findings: The incubator-hatched goslings followed Lorenz everywhere and showed no recognition of their natural mother. The goslings hatched by the mother followed her. This process, which Lorenz called imprinting, occurred within a brief, biologically determined critical period (approximately 13–16 hours after hatching).

Key principles:

• Imprinting is innate (genetically programmed).
• It occurs during a critical period.
• It is irreversible — once imprinted, the attachment cannot be reversed.
• Imprinting affects later mate preference — Lorenz found that imprinted birds later directed sexual behaviour towards humans.

Harlow (1958, 1962) — Contact comfort in rhesus monkeys:

Procedure: Harlow separated infant rhesus monkeys from their mothers at birth and raised them with two surrogate “mothers”: one made of wire mesh that provided milk, and one covered in soft cloth that provided no food.

Findings: The monkeys overwhelmingly preferred the cloth mother, spending most of their time clinging to it and only going to the wire mother to feed. When frightened, they always sought the cloth mother. Monkeys raised without a real mother showed severe social and emotional disturbance as adults, including difficulty mating and neglecting or abusing their own offspring.

Significance: Challenged the cupboard love theory (the idea that attachment forms because the caregiver provides food). Demonstrated that contact comfort is more important than feeding in the formation of attachment.

Human Attachment

Ainsworth et al. (1978) — The Strange Situation:

A controlled observation procedure designed to measure the quality of attachment between a child (12–18 months) and their caregiver. It takes place in a controlled room with a two-way mirror and consists of eight episodes:

1. Parent and child enter the room.
2. Child explores while parent sits quietly.
3. Stranger enters and talks to the parent.
4. Parent leaves; stranger interacts with the child.
5. Parent returns; stranger leaves (reunion).
6. Parent leaves the child alone.
7. Stranger returns.
8. Parent returns; stranger leaves (second reunion).

Observers record the child’s response to separation and reunion, their exploration, and their response to the stranger.

Three attachment types identified:

Type	Separation anxiety	Reunion behaviour	Stranger anxiety	Proportion (US sample)
Secure (Type B)	Distressed when parent leaves	Seeks contact and is readily comforted	Wary but can be comforted by parent	~66%
Insecure-avoidant (Type A)	Shows little distress	Avoids or ignores parent on return	Indifferent to stranger	~22%
Insecure-resistant (Type C)	Very distressed	Seeks and rejects contact (ambivalent)	Very fearful, cannot be comforted	~12%

Main and Solomon (1986, 1990) later identified a fourth type:

• Insecure-disorganised (Type D): The child shows no consistent pattern. They may freeze, rock, or display contradictory behaviours. Associated with frightening or abusive caregiving.

Cultural variations: Van Ijzendoorn and Kroonenberg (1988) conducted a meta-analysis of 32 Strange Situation studies across 8 countries. Secure attachment was the most common type in all cultures. However, the proportion of insecure-avoidant and insecure-resistant attachments varied across cultures — German samples showed higher rates of insecure-avoidant attachment, while Japanese samples showed higher rates of insecure-resistant attachment. This may reflect cultural differences in child-rearing practices rather than differences in attachment quality.

Bowlby’s Evolutionary Theory of Attachment

John Bowlby (1969, 1980) proposed that attachment is an innate, evolutionary adaptation that enhances survival.

Key principles:

1. Adaptive: Attachment is biologically programmed because it enhances survival. Infants who stay close to their caregivers are protected from predators and provided with food.

2. Social releasers: Infants are born with innate behaviours that elicit caregiving from adults — smiling, cooing, grasping, crying. These “social releasers” activate the adult’s innate caregiving response.

3. Critical period: Bowlby proposed that attachment must form within a critical period (approximately birth to 2.5 years). If no attachment forms during this time, the child will suffer irreversible long-term consequences. (Later research suggests a sensitive period is more accurate — attachment can form after this window, but it is more difficult.)

4. Monotropy: Bowlby argued that the child forms one special, primary attachment (in most cases to the mother), which is qualitatively different from other attachments. This primary attachment figure provides a secure base from which the child explores the world.

5. Internal working model: The child’s first attachment creates a template (internal working model) for all future relationships. A secure attachment leads to expectations that others are trustworthy and supportive, while an insecure attachment leads to expectations that others are unreliable or rejecting.

6. Continuity hypothesis: There is a consistent link between early attachment experiences and later emotional and social development. Securely attached children tend to have better relationships in adulthood; insecurely attached children tend to have more difficulties.

Evidence for the continuity hypothesis: Hazan and Shaver (1987) found that adults classified as securely attached in romantic relationships were more likely to describe their childhood relationships with parents as warm and supportive. Insecurely attached adults reported more jealousy, fear of closeness, and belief that romantic love is rare.

Deprivation and Institutionalisation

Bowlby’s Maternal Deprivation Hypothesis (1951):

Bowlby argued that prolonged separation from the primary attachment figure during the critical period would have irreversible negative consequences for the child’s emotional and social development, including affectionless psychopathy (the inability to feel guilt or form meaningful relationships).

The 44 Thieves Study (Bowlby, 1944):

Procedure: Bowlby identified 88 children attending a child guidance clinic. 44 were juvenile thieves; 44 were a control group (referred for emotional problems but not theft). Bowlby assessed whether the children had experienced prolonged early separation from their mothers.

Findings: 17 of the 44 thieves (39%) had experienced prolonged early separation, compared to only 2 of the 44 non-thieves (5%). Among the thieves, 14 were classified as “affectionless psychopaths,” and 12 of these 14 had experienced prolonged separation.

Evaluation: This study supports the maternal deprivation hypothesis. However, it is retrospective (relying on historical records and memory), and Bowlby may have been biased in his assessment of affectionless psychopathy (investigator effects). The correlation between separation and delinquency does not prove causation.

Romanian Orphan Studies:

After the fall of the Ceaușescu regime in Romania in 1989, thousands of children were found living in severely deprived institutional conditions. This provided a unique natural experiment.

Rutter et al. (2011) — English and Romanian Adoptees (ERA) study:

Procedure: 165 Romanian orphans adopted by British families were assessed at various ages (4, 6, 11, 15, and into adulthood). They were compared with a control group of 52 British children adopted before 6 months. The Romanian orphans were grouped by age of adoption: under 6 months, 6–24 months, and 24–42 months.

Findings:

• Children adopted before 6 months showed good recovery, with attachment outcomes similar to the British control group.
• Children adopted after 6 months showed higher rates of disinhibited attachment (indiscriminate friendliness, lack of appropriate social wariness), attention-seeking behaviour, and problems with peer relationships.
• However, even those adopted later showed significant developmental catch-up, particularly in cognitive development.

Significance: Demonstrates that the effects of institutionalisation can be partially reversed by early intervention, but the likelihood of recovery depends on the timing — supporting the concept of a sensitive (not critical) period. Disinhibited attachment is a particularly persistent consequence of early institutional care.

Zeanah et al. (2005): Studied 95 Romanian children (12–31 months) who had spent most of their lives in institutions. Compared with a control group who had never been institutionalised. Found that 65.4% of the institutionalised group showed disorganised attachment, compared to 24.1% of the control group.

Influence of Early Attachment on Later Relationships

Childhood friendships: Kerns (1994) found that securely attached children have more positive friendships, while insecurely attached children tend to have more conflict and less closeness in friendships.

Romantic relationships: Hazan and Shaver (1987) found that attachment type predicted adult romantic relationship quality. Secure adults had longer, more trusting relationships; avoidant adults feared intimacy; anxious-resistant adults were obsessed with their partners and experienced highs and lows.

Parenting: Bailey et al. (2007) found that 75% of mothers classified as securely attached had securely attached babies, supporting the intergenerational transmission of attachment through the internal working model.

Adult attachment type and mental health: Mikulincer and Shaver (2007) found that insecure attachment is associated with higher rates of anxiety, depression, and personality disorders in adulthood.

Key Studies

Study	Researcher(s)	Year	Method	Key Findings	Evaluation
Imprinting	Lorenz	1935	Observation (animals)	Goslings imprinted on the first moving object they saw; critical period	Ethical concerns (manipulation of animals); cannot generalise from geese to humans
Contact comfort	Harlow	1958	Lab experiment (animals)	Monkeys preferred cloth mother over wire mother providing food	Challenged cupboard love theory; severe animal cruelty; limited generalisability to humans
Strange Situation	Ainsworth	1978	Controlled observation	Identified three attachment types (secure, avoidant, resistant)	High reliability; culturally biased; low ecological validity
44 thieves	Bowlby	1944	Retrospective case comparison	39% of thieves had early prolonged separation	Supports deprivation hypothesis; retrospective; investigator bias
Maternal deprivation	Bowlby	1951	Literature review / clinical observations	Prolonged early separation leads to affectionless psychopathy	Influential policy impact; based on limited evidence; ignores other caregiver bonds
Romanian orphans	Rutter et al.	2011	Longitudinal study	Earlier adoption led to better outcomes; sensitive period supported	High ecological validity; ethical research; cannot control all variables
Meta-analysis of attachment	Van Ijzendoorn & Kroonenberg	1988	Meta-analysis	Secure attachment most common across all cultures; cultural variation in insecure types	Large sample across cultures; Strange Situation may not be valid for all cultures
Love quiz	Hazan & Shaver	1987	Self-report (questionnaire)	Adult attachment type linked to childhood attachment experiences	Large sample; correlational; self-report bias; retrospective

Key Terminology

Term	Definition
Attachment	A deep, enduring emotional bond between two people
Imprinting	An innate, biologically programmed response to follow the first moving object seen after birth
Critical period	A biologically determined time window during which attachment must form for normal development
Sensitive period	A time when a child is most receptive to forming attachments, but it is still possible afterwards
Monotropy	Bowlby’s idea that one primary attachment is most important and qualitatively different from others
Social releasers	Innate infant behaviours (smiling, crying, cooing) that elicit caregiving from adults
Internal working model	A mental template of expectations about relationships, based on early attachment experiences
Continuity hypothesis	The idea that early attachment experiences predict later relationship quality
Secure base	The primary attachment figure who provides a safe base from which the child explores
Strange Situation	A controlled observation procedure to assess attachment type in infants
Secure attachment	A healthy attachment characterised by distress at separation and comfort at reunion
Insecure-avoidant	An attachment type where the child avoids or ignores the caregiver
Insecure-resistant	An attachment type where the child is clingy but rejects the caregiver at reunion
Disinhibited attachment	Indiscriminate friendliness and lack of social wariness, associated with institutionalisation
Maternal deprivation	Prolonged separation from the primary attachment figure during the critical/sensitive period
Affectionless psychopathy	The inability to feel guilt or form meaningful emotional relationships

Evaluation Points

Strengths

• Lorenz and Harlow provide clear evidence that attachment has a biological basis and that contact comfort is more important than feeding. Harlow’s findings directly challenged the behaviourist cupboard love theory.
• The Strange Situation is a reliable and well-validated measure. Ainsworth et al. found inter-observer reliability of 0.94, and the procedure has been replicated consistently across many studies.
• Bowlby’s theory is supported by a wide range of evidence, including animal studies, the 44 thieves study, Romanian orphan research, and the continuity hypothesis (Hazan and Shaver). It has also been highly influential in shaping childcare policy (e.g., hospital visiting policies for children).
• Romanian orphan studies have high ecological validity, studying real deprivation in natural conditions. They demonstrate that recovery is possible, particularly with early intervention.

Limitations

• Animal studies raise issues of generalisability — attachment processes in geese and monkeys may not be the same as in humans. They also raise significant ethical concerns about the treatment of animals.
• The Strange Situation may not be culturally universal. Takahashi (1990) found that Japanese infants showed high rates of insecure-resistant attachment, likely because Japanese mothers rarely leave their children alone, making the separation episodes unusually stressful rather than indicative of genuine attachment insecurity.
• Monotropy is contested. Many cultures involve multiple caregivers (e.g., the Efe people of Zaire, where children are cared for communally). Thomas (1998) argued that children can form multiple attachments of equal importance.
• The continuity hypothesis is based largely on correlational and retrospective data. Hazan and Shaver’s study relied on adults recalling their childhood attachments, which is subject to memory distortion. Temperament may be an alternative explanation — a child’s innate temperament could influence both attachment type and later relationship quality.
• Deterministic: Bowlby’s theory implies that early attachment irrevocably determines later outcomes. This is overly deterministic, as the Romanian orphan studies show that recovery is possible, and many people with insecure attachments go on to have successful relationships.

Methodology

Attachment research uses a range of methods:

• Controlled observations (Strange Situation): High reliability and replicability but artificial setting.
• Longitudinal studies (ERA study, Rutter): Track development over time, allowing causal inferences, but time-consuming and subject to attrition.
• Self-report questionnaires (Hazan and Shaver): Efficient for large samples but subject to social desirability and retrospective bias.
• Animal experiments (Lorenz, Harlow): Allow manipulation of variables but raise ethical and generalisability concerns.
• Meta-analysis (van Ijzendoorn and Kroonenberg): Combine data from many studies for a more comprehensive picture, but rely on the quality of the included studies.

Common Pitfalls

1. Confusing deprivation and privation: Deprivation is the loss of an attachment that has already formed (e.g., through separation). Privation is the failure to form any attachment at all (e.g., institutionalised children who never had a primary caregiver). The effects and recovery prospects differ.
2. Confusing critical period and sensitive period: Bowlby originally used “critical period” (implying no recovery is possible after the window closes). Research from Romanian orphan studies supports the concept of a “sensitive period” — recovery is harder but not impossible. Always use the correct term.
3. Overstating monotropy: Bowlby’s emphasis on the mother as the sole attachment figure is culturally biased and outdated. Modern research recognises multiple attachment figures. Do not assume the mother is always the primary attachment figure.

Worked Examples

Example 1: 16-Mark Essay

Question: Discuss Bowlby’s evolutionary theory of attachment. Refer to evidence in your answer. [16 marks]

Model Answer:

Bowlby’s evolutionary theory of attachment (1969) proposes that attachment is an innate, adaptive behaviour that has evolved because it enhances infant survival. According to Bowlby, infants are born with an innate drive to form an attachment with their primary caregiver, who provides protection, nourishment, and a secure base from which to explore.

A central concept is that of social releasers. Bowlby argued that infants are born with innate behaviours — smiling, cooing, crying, and grasping — that elicit caregiving responses from adults. These social releasers activate the adult’s innate caregiving system, creating a reciprocal attachment process. This explains why adults generally find infants appealing and feel compelled to respond to their cries.

Bowlby proposed that attachment must form within a critical period, approximately the first two and a half years of life. If no attachment forms during this time, the child will suffer irreversible consequences for their emotional and social development. This was based on his maternal deprivation hypothesis, supported by the 44 thieves study (1944), in which 39% of juvenile thieves had experienced prolonged early separation from their mothers, compared to only 5% of the control group.

The concept of monotropy is central to Bowlby’s theory. He argued that the child forms one special, primary attachment that is qualitatively different from all others. This primary attachment relationship provides the basis for the internal working model — a mental template of expectations about relationships that guides all future social interactions. A secure attachment leads to the expectation that others are trustworthy and supportive, while an insecure attachment leads to expectations of rejection or unreliability.

The internal working model forms the basis of the continuity hypothesis, which states that there is a consistent link between early attachment and later relationship quality. Hazan and Shaver (1987) provided support for this in their “love quiz” study, finding that adults who described their childhood relationships as warm and supportive were more likely to be securely attached in their romantic relationships, while those who described distant childhood relationships were more likely to be avoidant.

A significant strength of Bowlby’s theory is its strong evolutionary basis. The idea that attachment is adaptive and enhances survival is consistent with the principles of natural selection. Lorenz’s (1935) research on imprinting in geese supports the concept of an innate, biologically programmed attachment process. Harlow’s (1958) research on rhesus monkeys further supports the idea that attachment is not merely based on feeding but involves a deeper, innate emotional bond.

Furthermore, Bowlby’s theory has had enormous practical impact. It led to changes in hospital policies, allowing parents to stay with their children during hospitalisation. It also influenced the development of childcare practices, such as key workers in nurseries who provide a consistent attachment figure.

However, Bowlby’s concept of monotropy has been heavily criticised. Many cultures involve multiple caregivers from birth. For example, the Efe people of Zaire share childcare communally, with infants being passed between multiple caregivers. This contradicts the idea that one primary attachment is essential. Thomas (1998) argued that children can and do form multiple attachments of equal importance, and that there is no evidence that one primary attachment is qualitatively different.

The critical period has also been challenged. Rutter’s English and Romanian Adoptees study (2011) found that children adopted after the critical period still showed significant recovery, particularly in cognitive development. This supports the concept of a sensitive period rather than a strict critical period, suggesting that Bowlby was overly pessimistic about the possibility of recovery.

Additionally, the continuity hypothesis is supported largely by correlational data. Temperament may offer an alternative explanation — a child’s innate temperament (e.g., being easy-going or difficult) could influence both attachment type and later relationship quality, without early attachment being the cause of later outcomes. Kagan (1984) argued that temperament is a better predictor of attachment type than caregiving quality.

In conclusion, Bowlby’s evolutionary theory provides a compelling and influential account of attachment, supported by animal research, clinical evidence, and the continuity hypothesis. However, its emphasis on monotropy and the critical period has been challenged, and the theory may be culturally biased and overly deterministic.

Example 2: 16-Mark Essay

Question: Discuss research into the effects of institutionalisation. Refer to Romanian orphan studies in your answer. [16 marks]

Model Answer:

Institutionalisation refers to the placement of children in institutional care (such as orphanages) where they may experience a lack of consistent caregiving. Research into institutionalisation has been crucial for understanding the effects of early deprivation on development and for informing childcare policy.

The Romanian orphan studies, particularly Rutter et al.’s (2011) English and Romanian Adoptees (ERA) study, represent the most significant research in this area. After the fall of the Ceaușescu regime in 1989, over 100,000 Romanian children were found living in extremely deprived institutional conditions — shared carers, minimal interaction, poor nutrition, and little stimulation. Rutter and colleagues followed 165 Romanian orphans adopted by British families, comparing them with 52 British children adopted before 6 months of age.

The findings were striking. Children adopted before 6 months of age showed good recovery, with cognitive development and attachment outcomes similar to the British control group by age 4. This suggests that early adoption can largely mitigate the effects of institutionalisation. However, children adopted between 6 and 24 months, and those adopted between 24 and 42 months, showed higher rates of developmental difficulties, particularly disinhibited attachment — characterised by indiscriminate friendliness, a lack of appropriate social wariness, and attention-seeking behaviour.

Disinhibited attachment was the most persistent consequence of institutionalisation. Even at age 15, children adopted after 6 months continued to show higher rates of disinhibited attachment than those adopted earlier. However, cognitive development showed remarkable catch-up — by age 11, even those adopted after 24 months showed significant improvement in IQ, although they still lagged behind the control group.

A major strength of the Romanian orphan studies is their high ecological validity. Unlike laboratory studies, these children experienced real deprivation in real institutions, and their recovery occurred in real adoptive families. This makes the findings highly generalisable to other situations of institutional care. The longitudinal design (following participants over many years) also allowed researchers to track developmental trajectories and assess long-term outcomes.

Furthermore, the research has had significant practical applications. It has informed adoption policy, supporting the importance of early placement in adoptive families. Many countries have subsequently reduced the time children spend in institutional care before adoption, and some have prioritised foster care over institutional care.

However, the Romanian orphanages represented an extreme case of deprivation. The conditions were exceptionally poor, even by the standards of institutional care. This means the findings may not generalise to institutional care in other countries, where conditions may be less severe. The children also experienced multiple forms of deprivation — not just emotional but also nutritional, cognitive, and medical — making it difficult to isolate the specific effect of attachment deprivation.

Zeanah et al. (2005) provided complementary evidence from the Bucharest Early Intervention Project (BEIP), a randomised controlled trial assigning Romanian institutionalised children to either continued institutional care or high-quality foster care. Children placed in foster care showed significantly better attachment outcomes, with the greatest improvement in those placed before 20 months of age. This study has the added strength of experimental control (random allocation), strengthening the causal inference that institutional care causes poor attachment outcomes.

However, the BEIP also raised ethical concerns. Randomly assigning children to remain in institutional care when foster care was available could be considered unethical. The researchers argued that the foster care was provided as part of the study and would not have existed otherwise, but the ethical debate continues.

In conclusion, Romanian orphan studies have provided compelling evidence that institutionalisation has negative effects on attachment, cognitive development, and social behaviour. However, they also demonstrate that recovery is possible, particularly with early intervention, supporting the concept of a sensitive rather than critical period. The research has had significant practical impact on childcare and adoption policy worldwide.

Summary

Attachment research demonstrates the critical importance of early emotional bonds:

• Animal studies (Lorenz, Harlow) established the biological basis of attachment and the importance of contact comfort over feeding.
• The Strange Situation (Ainsworth) identified three main attachment types: secure, insecure-avoidant, and insecure-resistant.
• Bowlby’s evolutionary theory proposes that attachment is innate and adaptive, operating through social releasers, monotropy, and the internal working model.
• Deprivation and institutionalisation can have serious effects, but recovery is possible with early intervention (supporting a sensitive rather than critical period).
• Early attachment influences later relationships through the continuity hypothesis, although temperament and later experiences also play a role.