According to Mary Ainsworth (1974), the type and quality of attachment between mother
and child is largely dependent on the mother’s behaviour towards the child. She
suggested that mothers of securely attached infants tended to be more sensitive to
the child’s needs - more responsive, more co-operative and more accessible than mothers
of either of the anxious types. These mothers provide a warm and close physical contact,
especially when the infant is distressed. Ainsworth’s concept is called the Caregiver
Sensitivity Hypothesis.
Ainsworth proposed that the caregivers of anxious-resistant infants were interested
in them but misunderstood the infant’s behaviour. Of particular importance, these
caregivers tended to be inconsistent in the way they treated their children. As a
result, the infant is unable to rely on the caregiver’s emotional support.
Ainsworth reported that caregivers of anxious-avoidant infants were often impatient
and/or uninterested, often rejecting them and tending to be self-centred and rigid
in their behaviour. However, some caregivers of avoidant infants acted in a suffocating
way, interacting with their infants even when the infants did not want any interaction.
A powerful comment on the impact of the mother’s state on the formation of attachments
came from D M Teti, D M Gelfand, D S Messinger & R Isabella (1995) who showed a significant
association between maternal Depression and infant attachment insecurity. They found
that children without coherent attachment strategies tended to have more chronically
impaired mothers than did securely-attached children.
In a 1989 study of neglected and badly-treated infants V Carlson, D Cicchetti, D
Barnett & K Braunwald found that about 80% of them matched Type D disorganised criteria.
They concluded that the highly-stressful and inconsistent regime in an abusive home
may interfere with the organisation of an effective attachment system.
Michael Lamb, Ross Thompson, William Gardner & Eric Charnov (1985) surveyed research
across the United States and came to the view that sensitive and responsive mothering
does indeed lead to secure attachment. Alan Sroufe & June Fleeson (1986) argued that
the common factor in the lives of securely-attached infants seems to be the contingent
responsiveness of the parents to the infant’s needs.
Mary Main, Nancy Kaplan & Jude Cassidy (1985) explored the relationship between the
mother’s behaviour and type of attachment and found that mothers who themselves had
had satisfactory attachment experiences would be more likely to foster secure attachments
in their own children.
Using a procedure known as the Adult Attachment Interview, Main, Kaplan & Cassidy
classified parents according to their recollections of their own attachment experiences.
Autonomous (secure) adults, who discussed their own childhood experiences openly,
were more likely to have children identified as securely attached in a Strange Situation.
Dismissing and pre-occupied adults, however, tended to have children identified as
insecure in the Strange Situation.
According to Main, Kaplan & Cassidy, a mother’s own experiences affects how she interacts
with her child and the type of relationship established between the mother-child
pair. This can be linked to John Bowlby’s concept of the internal working model.
The
Temperament Hypothesis
An alternative explanation for the correlation between early attachment types and
later development is the Temperament Hypothesis of Jerome Kagan (1984). Kagan argued
that this can be explained by the infant’s innate temperament. J J Campos, K C Barrett,
M E Lambe, H H Goldsmith & C Sternberg (1983) point out that differences in maternal
behaviour may be due in part to the characteristics of the child. Effectively, some
people are born good at forming relationships and some aren’t.a mother may appear
to be insensitive because her child is unresponsive.
Infants can be even-tempered or irritable, responsive or withdrawn, active or inactive,
soothed easily or with difficulty. Alexander Thomas & Stella Chess, in a 1977 longitudinal
study of 138 infants in New York City, classified 40% of the infants as having ‘easy’
temperaments, 10% were ‘difficult’, 15% ‘slow to warm up’ and 35% were ‘mixed’. Earlier,
Thomas, Chess & Herbert Birch (1970), in a study of 141 children from birth to 14
years, had shown that temperamental categorisations assigned shortly after birth
sustained by and large throughout their childhood. Since many aspects of temperament
are evident during the first weeks of life and continue into adulthood, Thomas &
Chess assumed they are largely inborn or innate.
Kagan’s views built on the earlier work of E E Maccoby (1980) who argued that ‘normal’
infants tend to take the initiative in inviting a maternal response but others do
not. Thus, responsive mothers may be at least partly a consequence of the child’s
characteristics - ie: a mother may become insensitive because her child is unresponsive.
An interesting bit of evidence to support the Temperament Hypothesis came from a
study in Japan by Kazuo Miyake, S J Chen & Campos (1985) which found that newborn
babies who became upset when their feeding was interrupted were more likely to be
classified as insecurely attached at 12 months.
Jay Belsky & Michael Rovine (1987) gave some support to the Temperament Hypothesis
in reporting that newborns who showing signs of behavioural instability - eg: tremors
or shaking or being startled easily - tended to be less securely attached to their
mother. They also tended to fall into certain sub-types associated with distress
in the Strange Situation.
Kagan also argued that the Caregiver Sensitivity Hypothesis placed too much emphasis
on the role played by the caregiver in the development of attachment and ignored
the part played by the infant’s temperament.
However, Kevin Durkin (1995) has pointed out that infants’ temperament, as assessed
by their parents, is not usually associated with their attachment type, as determined
by the Strange Situation. Earlier B E Vaughn, G B Lefever, R Seifer & P Barglow (1989)
had come to a similar conclusion.
The
Bigger Picture
Belsky & Rovine postulated that it is the interaction between relationship (caregiver
sensitivity) and temperament which influences the development of attachment. Certainly
Gottfried Spangler (1990), in a study of German mothers, found that their perceptions
of their infant’s temperament influenced their responsiveness.
Changes in family circumstances - such as the mother taking a job outside the home
- can also affect the development and stability of attachment types. Ross Thompson,
Michael Lamb & David Estes (1982) found 47% of middle-class mothers and infants in
their sample changed their type of attachment relationship between
