Stages of Attachment

Attachment is the emotional bond that forms between an infant and the people who care for it. John Bowlby and Schaffer & Emerson each mapped how this bond develops, from a newborn with no preference for any one caregiver to a toddler with several lasting attachments.

Stages	John Bowlby	Schaffer and Emerson
1	Pre-attachment (Birth to 6 Weeks): The infant does not discriminate between caregivers.	Asocial Stage (0-6 weeks): Infant behavior is directed at anyone or anything with a positive reaction, such as a smile.
2	Attachment-in-the-Making (6 Weeks to 7 Months): Prefers familiar people but accepts care from anyone.	Indiscriminate Attachment (6 weeks-7 months): Infants prefer people to inanimate objects, but not one person specifically.
3	Clear-Cut Attachment (7-24 Months): Specific attachment to the primary caregiver(s) develops.	Specific Attachment (7-9 months): Infants prefer particular caregivers and look to them for comfort, security, protection, and reassurance in stressful situations.
4	Formation of Reciprocal Relationships (24 months +): Children will acquire language skills and better understand the pattern of carers leaving and returning.	Multiple Attachments (10 months and onwards): Infants become increasingly independent and form attachments with multiple people in their lives who respond sensitively to them, including grandparents, siblings, or neighbors.

Schaffer & Emerson’s (1964)

Rudolph Schaffer and Peggy Emerson (1964) studied 60 babies (31 boys and 29 girls) at monthly intervals during their first year, with a follow-up visit at 18 months. This is known as a longitudinal study.

The children were all studied in their own homes, and a regular pattern was identified in the development of attachment. Researchers observed the babies’ interactions with their carers and interviewed the carers at each visit.

The mother kept a diary, and the following measures were recorded:

• Stranger Anxiety: response to the arrival of a stranger.
• Separation Anxiety: distress level when separated from a carer, and how much comfort was needed on return.
• Social Referencing: how much a child looks to a carer to check how to respond to something new (a secure base).

They discovered that baby’s attachments develop in the following sequence:

1. Asocial (0 – 6 weeks):

   Very young infants are asocial in that many kinds of stimuli, both social and non-social, produce a favorable reaction, such as a smile.
   
   During the asocial stage, the infant’s behavior does not appear specifically directed towards primary caregivers over others. The infant may respond to humans and objects similarly, showing generalized responses like crying, smiling, and looking at faces or objects.
   
   Although termed “asocial,” it doesn’t imply that the baby has no interest in social interaction. Still, rather than undifferentiated interactions, they do not yet show a strong preference for any one person or display an understanding of social behavior. This is the period before specific, strong social bonds are formed.
   
   Remember that despite its name, the asocial stage is important for social development. The interactions during this stage lay the groundwork for later stages of attachment.

2. Indiscriminate Attachments (6 weeks to 7 months):

   During this stage, infants start to display more social behavior and begin to show a preference for social stimuli, preferring human interaction over inanimate objects. They smile, babble, and reach out more to people rather than objects.
   
   Infants indiscriminately enjoy human company; most babies respond equally to any caregiver. They get upset when an individual ceases to interact with them.
   
   The infants are sociable and “indiscriminate” in their attachment because they do not show a strong preference for a single caregiver over others. They recognize and respond to more people, but they do not yet differentiate significantly between familiar and unfamiliar adults.
   
   From 3 months, infants smile more at familiar faces and can be easily comfortable by a regular caregiver.

3. Specific Attachment (7 – 9 months):

   Special preference for a single attachment figure.  The baby looks to particular people for security, comfort, and protection.  It shows fear of strangers (stranger fear) and unhappiness when separated from a special person (separation anxiety).
   
   Some babies show stranger fear and separation anxiety much more frequently and intensely than others, but they are evidence that the baby has formed an attachment.  This has usually developed by one year of age.
   
   The study’s results indicated that attachments were most likely to form with those who responded accurately to the baby’s signals, not the person they spent more time with. Schaffer and Emerson called this sensitive responsiveness.
   
   In 39% of cases, the person who did the most physical caretaking (usually the mother) was not the baby’s primary attachment figure. The strongest predictor was sensitive responsiveness to the baby’s signals, not who fed them or spent the most time with them.
   
   Intensely attached infants had mothers who responded quickly to their demands and, interacted with their child. Infants who were weakly attached had mothers who failed to interact.
   
   The most important fact in forming attachments is not who feeds and changes the child but who plays and communicates with him or her. Therefore, sensitive responsiveness to the baby’s signals appeared to be the key to the attachment.

4. Multiple Attachment (10 months and onwards):

   The baby becomes increasingly independent and forms several attachments. By 18 months, the majority of infants have formed multiple attachments.
   
   By around 7 months, 29% of infants had already formed several attachments, and 10% had five or more. By 10 months, 59% had more than one attachment. By 18 months, 87% had formed multiple attachments.
   
   The primary attachment was not always the mother. At 18 months, only about half of infants were most strongly attached to their mother, around a third to their father, and about 17% equally to both parents. This directly challenges Bowlby’s idea that one mother-figure is uniquely primary.

Together, these findings led Schaffer and Emerson to reject “cupboard love”: babies attach to whoever responds sensitively, not whoever feeds them. As Schaffer later put it, the infant is not a passive baby who simply “lives to eat,” but an active one who “eats to live” (Schaffer, 1971).

The study also has real methodological strengths that balance these limitations:

• Longitudinal Design: The same infants were tracked over time, so age differences reflect real developmental change rather than differences between groups of different children.
• Ecological Validity: Observations took place in the infants’ own homes during everyday routines, making the findings more likely to reflect how attachment develops in real life.
• Theoretical Value: The study produced a testable model of sensitive responsiveness that has shaped decades of further research and has broad applied relevance for childcare and parenting.

Critical Evaluation

Several limitations affect how far the findings generalise:

• Population Validity: The 60 infants came from Glasgow, mostly working-class families, so the sample cannot represent other social classes or cultures.
• Diary Reliability: Data came largely from mothers’ diaries and reports, which may involve social desirability bias and under-reporting of distress, since diaries were not kept by trained, objective observers.
• Temperament Confound: The link between caregiver sensitivity and attachment is correlational; an “easy,” sociable infant may simply be easier to respond to and quicker to attach regardless of caregiving.
• Historical Validity: The study was conducted in 1960s Glasgow, when fathers rarely did routine childcare, so its timetable may not generalise to modern parenting.
• Culture-Bound Timetable: The stage sequence reflects a Western, individualist family structure; in many collectivist cultures, infants form multiple attachments from birth (Tronick et al., 1992).
• Asocial Stage Is Hard to Study: A newborn cannot yet show clear preferences, so what the study reports about the earliest weeks rests more on adult interpretation than on direct evidence.

Contemporary Research

Modern research has tested Schaffer and Emerson’s central claims with far larger samples than a 60-infant diary study could offer. The strongest evidence comes from meta-analyses that pool many studies of infants’ attachments to both parents.

• Aim: To test how an infant’s attachment network (its bonds to both parents considered together), not the mother-child bond alone, predicts later internalising and externalising behaviour problems.
• Method: Researchers pooled individual-level data from 9 studies covering 1,097 children who had been assessed for attachment to both their mother and father, mostly using the Strange Situation.
• Results: Children insecurely attached to one or both parents had more internalising problems. Children disorganised with both parents had more externalising problems, and mother-child and father-child attachment mattered in similar ways.
• Conclusion: A child’s whole attachment network predicts adjustment better than the mother-child bond alone, and the father is a developmentally important attachment figure in his own right.

A further meta-analysis by Pinquart (2022) compared attachment security to mothers and fathers across 95 studies. The gap was small. A standardised mean difference of about g = -0.07 put fathers only slightly behind mothers. Security to the two parents was also moderately correlated (r = 0.32).

This matters.

It supports Schaffer and Emerson’s claim that fathers are genuine attachment figures, not weaker copies of the maternal bond. It also shows a child’s bonds to each parent are partly independent, rather than one simply mirroring the other.

Cross-cultural evidence points the same way. Among the Efe foragers of the Democratic Republic of Congo, infants are cared for by many adults from birth. They form several attachments at once, without passing through a single-carer bond first (Tronick et al., 1992).

The exact sequence may not be universal.

Still, the wider point holds: infants readily form multiple attachments, whatever order those bonds arrive in.

A newer, smaller study adds a caution. Paquette and colleagues (2024) followed 143 two-parent families, classifying each child’s attachment to both mother and father at 15 months and again at 45 months. They found no significant stability in these classifications over time, for either parent or the pair together.

Bowlby’s Phases of Attachment?

John Bowlby, a British psychiatrist, and psychoanalyst, developed attachment theory, which focuses on the importance of a secure and trusting mother-infant bond on development and wellbeing.

He identified four phases of attachment.

These stages reflect a process where each step builds upon the last, leading to a strong, secure attachment under ideal circumstances.

Bowlby believed that these attachments serve a crucial evolutionary function: they increase the infant’s chance of survival by ensuring its safety and care.

1. Pre-attachment phase (Birth to 6 Weeks):

   In this stage, infants show innate behaviors (like crying and smiling) that help attract caregiver attention and response but do not prefer a specific caregiver.
   
   They accept care and comfort from any adult and do not yet understand that they are separate individuals from their caregivers.
   
   Even so, the raw materials of sociability are already there. Behaviours like smiling and being soothed by touch lay the groundwork for the preferences that appear in later stages.

2. Attachment-in-the-Making (6 Weeks to 7 Months):

   During this stage, infants develop a sense of trust in the people who respond to their needs.
   
   They start to distinguish between familiar and unfamiliar people, preferring familiar people, but they still accept care from strangers. They don’t yet protest when separated from a parent.

3. Clear-Cut Attachment (7 Months to 24 Months):

   Specific attachment to the primary caregiver(s) develops at this stage.
   
   Infants seek regular contact with their caregivers and may show distress upon separation, demonstrating behaviors like “clinging” and “following.”
   
   This stage also includes the development of “stranger anxiety” and “separation anxiety.”
   
   The onset of stranger anxiety around 7-9 months reflects cognitive advances enabling infants to distinguish familiar from unfamiliar people.
   
   While strangers provoke uncertainty, repeated comforting interactions with a sensitive caregiver provide the infant with the confidence to overcome this wariness.
   
   By age 2, stranger anxiety typically resolves as the child becomes more socially adept.

4. Formation of a Reciprocal Relationship (24 Months and Onwards):

   As language develops, toddlers understand their parents’ actions and motivations and can negotiate with them.
   
   They can now tolerate separations with less distress and use mental representations of their caregivers for comfort.

Critical Evaluation

• Limited Discrimination at Birth: Bowlby proposed that newborns cannot yet discriminate between caregivers and will accept care from whoever provides it, relying mainly on crying to get it.
• Early Recognition Abilities: Babies can recognise their mother’s voice and smell soon after birth (Kisilevsky et al., 2003; Porter & Winberg, 1999).
• Mother’s Presence Not Required: This early recognition does not mean only the mother can soothe a newborn; many caregivers can calm a distressed baby.
• Early Self-Regulation: A newborn’s cry-and-be-soothed pattern with caregivers is an early attempt at self-regulation.
• Growing Behavioural Repertoire: Over the first months, babies develop more attachment behaviours beyond crying.

Real-World Applications

The study’s findings shape real practice too. Sensitive responsiveness builds the bond, and infants readily form several attachments. Both ideas show up well beyond a developmental timetable.

Childcare and the Key Person

UK early-years settings must assign each child a key person. This is a named practitioner who takes primary responsibility for that child’s day-to-day care.

The goal is a consistent, responsive relationship with the child.

A caregiver becomes an attachment figure through responsive, stimulating interaction. Kinship and feeding duties matter far less than how someone actually responds.

A key person can become a real secondary attachment figure, alongside the parents themselves.

This can ease the separation anxiety a new nursery place often triggers, once a child has reached the specific-attachment stage. Continuity of staffing matters more than ratios alone.

A changing cast of carers gives an infant no single responsive figure to attach to. That undermines the very mechanism Schaffer and Emerson identified.

Fostering and Adoption

Attachment does not require a biological link. Sensitive responding is what counts.

Because of this, fostering and adoption practice expects a new carer to become a real attachment figure relatively quickly. The placement just needs to offer consistent, responsive care.

This follows directly from the “cupboard love” challenge.

An adoptive or foster parent who did not carry the child, breastfeed it, or share its genes is not disqualified from becoming its primary attachment figure. Biology and duration matter less than how the carer responds.

Disrupted or repeatedly changed placements are a particular concern for this same reason.

They deny a child the sustained, responsive relationship the stage model shows attachment depends on. Placement stability is protective in its own right.

Paediatric and Hospital Care

Hospitals once routinely restricted parental visiting.

This rested on a mistaken assumption. A young child’s distress at a parent leaving was treated as a passing nuisance, not a sign of a real attachment relationship under threat (Bowlby, 1973).

Modern practice reads that distress differently.

Separation anxiety is now understood as the behavioural signature of an already-formed attachment, timed by the Glasgow data to emerge around 7-9 months. By the toddler years a child typically has several attachment figures, not one.

This has fed into flexible visiting and “rooming-in” today.

Hospitals now make room for whichever attachment figure the child actually has, mother, father, or another primary carer, rather than assuming only the mother’s presence matters. This also helps clinicians spot an age-appropriate response.

Fathers and Parental Leave

Infants attach to non-caretaking fathers just as readily as to caretaking mothers.

Modern evidence, covered above in Contemporary Research, shows mother-child and father-child attachment carry comparable developmental weight and near-equivalent security (Dagan et al., 2021; Pinquart, 2022).

This has practical weight.

The finding has been used to support parental-leave and workplace policies. These treat paternal involvement in early caregiving as developmentally valuable in its own right, not a secondary substitute for maternal care.

The wider point reaches beyond fathers, too.

Attachment depends on responsive interaction, not on feeding. Because of this, anyone who interacts sensitively with an infant can become a genuine attachment figure: a grandparent, a childminder, a day-nursery key worker.

Exclusive maternal care is simply not biologically mandatory.

Comparing Attachment Theories

Schaffer and Emerson’s stages are best understood alongside the other major frameworks for explaining and measuring infant attachment. Each asks a related but different question, and together they build a fuller picture of how attachment works.

Against Bowlby’s Monotropy

The study is usually read as evidence against strict monotropy, and rightly so.

Multiple attachments form quickly. The father is often a primary figure. Both facts are hard to square with one uniquely primary maternal bond (Bowlby, 1969; Rutter, 1981).

The relationship is more nuanced than a simple refutation, though.

Schaffer and Emerson also found that one attachment is usually strongest. This fits a softer reading of Bowlby’s own later position: several attachments exist within a hierarchy, topped by one figure.

What the study genuinely undermines is the identity and exclusivity of that primary figure. It need not be the mother, nor the feeder.

It coexists from very early on with a real network of other attachments. Read this way, the account complements rather than simply contradicts Bowlby’s theory.

Both agree attachment is biologically prepared and interaction-based. Schaffer and Emerson just replace a single dominant tie with a plural, responsiveness-determined network.

A Family Attachment Network

Some modern researchers go further than a hierarchy model.

They argue a child’s attachments are best modelled as a genuine family-level network. This differs from one primary bond plus subordinate others, or several independent one-to-one relationships.

A 2024 study followed 143 two-parent families.

It asked what best predicted their wellbeing.

The answer was not the security of any single mother- or father-child bond taken alone. It was whether the child was securely attached to both parents together, as a stable network (Paquette et al., 2024).

The pattern held across both wellbeing measures.

Children stably secure with both parents had parents reporting better mental health and marital satisfaction. The children themselves showed fewer externalising problems than children stably secure with only one parent.

This extends Schaffer and Emerson’s own multiple-attachment finding. But it shifts the question, from who is the strongest attachment, to how the whole configuration works together.

Ainsworth’s Strange Situation

Ainsworth’s Strange Situation is not a rival account of when attachments form.

It measures something different: the quality of an already-formed bond.

Schaffer and Emerson’s home-visiting design charted the timetable and targets of attachment.

It asked when a first bond appears, and to whom.

Ainsworth’s later, structured laboratory procedure instead classifies the quality of an already-formed attachment. It observes an infant’s response to brief separations and reunions with a caregiver.

The two are complementary, not competing.

Schaffer and Emerson’s “sensitive responsiveness” is precisely what Ainsworth’s programme went on to test as caregiver sensitivity, the central predictor of a secure Strange Situation classification (Ainsworth et al., 1978).

A full account of infant attachment draws on both. The Glasgow study answers when and to whom; the Strange Situation answers how secure.

References

Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. Basic Books.

Bowlby, J. (1973). Attachment and loss. Volume II. Separation, anxiety and anger. In Attachment and loss. volume II. Separation, anxiety and anger (pp. 429-p).

Bowlby, J. (1980). Loss: Sadness & depression. Attachment and loss (vol. 3); (International psycho-analytical library no.109). London: Hogarth Press.

Dagan, O., Schuengel, C., Verhage, M. L., van IJzendoorn, M. H., Sagi-Schwartz, A., Madigan, S., Duschinsky, R., Roisman, G. I., Bernard, K., Bakermans-Kranenburg, M., Bureau, J.-F., Volling, B. L., Wong, M. S., Colonnesi, C., Brown, G. L., Eiden, R. D., Fearon, R. M. P., Oosterman, M., Aviezer, O., … Cummings, E. M. (2021). Configurations of mother–child and father–child attachment relationships as predictors of internalizing and externalizing behavioral problems: An individual participant data (IPD) meta-analysis. New Directions for Child and Adolescent Development, 2021(180), 67–94. https://doi.org/10.1002/cad.20450

Kisilevsky, B. S., Hains, S. M., Lee, K., Xie, X., Huang, H., Ye, H. H., … & Wang, Z. (2003). Effects of experience on fetal voice recognition. Psychological science, 14(3), 220-224.

Paquette, D., Dubois-Comtois, K., Cyr, C., Lemelin, J.-P., Bacro, F., Couture, S., & Bigras, M. (2024). Early childhood attachment stability to mothers, fathers, and both parents as a network: Associations with parents’ well-being, marital relationship, and child behavior problems. Attachment & Human Development, 26(1), 66–94. https://doi.org/10.1080/14616734.2024.2338089

Pinquart, M. (2022). Attachment security to mothers and fathers: A meta-analysis on mean-level differences and correlations of behavioural measures. Infant and Child Development, 31(6), e2364. https://doi.org/10.1002/icd.2364

Porter, R., & Winberg, J. (1999). Unique salience of maternal breast odors for newborn infants. Neuroscience & Biobehavioral Reviews, 23(3), 439-449.

Schaffer, H. R., & Emerson, P. E. (1964). The development of social attachments in infancy. Monographs of the Society for Research in Child Development, 29(3), 1–77. https://doi.org/10.2307/1165727

Tronick, E. Z., Morelli, G. A., & Ivey, P. K. (1992). The Efe forager infant and toddler’s pattern of social relationships: Multiple and simultaneous. Developmental Psychology, 28(4), 568–577. https://doi.org/10.1037/0012-1649.28.4.568