• Children develop a hierarchy of attachments with
their various caregivers. For example, a child with
three different caregivers (mother, father and nanny)
will have a specific attachment relationship with each
caregiver based on how that specific caregiver responds
to the child in times when the child is physically hurt,
ill or emotionally upset; particularly, when frightened.
If the mother reacts in loving ways most of the time,
the child will develop an organized and secure
attachment with the mother. That same child could
develop an organized, insecure and avoidant
attachment with the father if the father reacts in
rejecting ways to the child’s distress most of the time.
That same child could develop a disorganized
attachment with the nanny if the nanny displays
atypical behaviours during interactions with the child
and has unresolved mourning or trauma.
• In situations with multiple foster placements, neglect
or institutionalization, children may develop disorders
of nonattachment (49).
• Reactive attachment disorder (RAD) is a special
problem. The diagnosis of RAD, whether using
criteria from the International Classification of
Diseases: Clinical Descriptions and Diagnostic
Guidelines (46) or Diagnostic and Statistical Manual of
Mental Disorders, 4th edition (47), was developed
without the benefit of data, and research evidence to
support its validity are still sparse (2). Zeanah and
his colleagues (48,49) criticized the criteria for RAD
as inadequate to describe children who have seriously
disturbed attachment relationships rather than no
attachment relationships. Another significant
problem with the psychiatric diagnosis of RAD is
that it suggests that the attachment difficulties lie
within the child (ie, it is the child who receives the
psychiatric diagnosis), when in fact, attachment
involves the relationship between a child and
caregiver. Finally, to my knowledge, there is no
convincing empirical evidence to suggest that RAD
is associated with any of the four types of attachment
(secure, avoidant, resistant and disorganized).
• Will letting an infant cry during the first six months of
life affect the attachment relationship between that
infant and the caregiver who lets the infant cry? Many
child protection workers and health and mental
professionals recommend that parents place a child safely
in a crib when frustrated or angry instead of shaking the
baby. Such a recommendation should continue to be
made; however, one should closely monitor how
frequently the parent needs to place the child in the crib
and not respond. It is also acceptable for a child to cry
when intrusive medical procedures need to be done to
save the life of a child, treat a sick infant or give
immunizations. Although, it may be advisable to have
the primary caregiver present and promptly hold and
comfort the infant. However, letting a baby cry because
it is ‘good for their lung development’ (as some parents
argue clinically), because it will ‘spoil’ the baby or
because the baby needs to find their own ways to selfsoothe
might not be advisable during the first six months
of life. Similarly, it is acceptable to let a baby cry during
the second six months of life when the crying is not
related to attachment (eg, when the child is not
physically hurt, ill or frightened/emotionally distressed).
Therefore, it is acceptable, from an attachment
perspective, to use the Ferber method (50) or another
sleep method, but only if the child does not have an ear
infection, teething, etc.
• During the first six months of life, promptly picking up a
baby who is crying is associated with four major outcomes
by the end of the first year of life. First, the baby cries
less. Second, the baby has learned to self-soothe. Third, if
the baby needs the caregiver to soothe him/her, the baby
will respond more promptly. And finally, the caregiver
who responded promptly and warmly most of the time
(not all the time; nobody can respond ideally all of the
time) to the baby’s cries, will have created secure,
organized attachment with all of the associated benefits.