provide. Emotional problems such as depression, economic stress, and marital conflict can interfere with sensitive and responsive parenting, be disruptive of secure attachments (see Belsky and Isabella, 1988; Thompson, 1999b; Waters, 1978), and constitute a significant source of instability over time in attachment security.
It also appears to be the case that atypical attachments are more common among atypical samples, including premature infants, children with Down syndrome, and children with autism (Atkinson et al., 1999; Capps et al., 1994). In particular, a significantly larger share of children at the extremes of reproductive risk or who have an identifiable developmental disability display disorganized or unclassifiable patterns of attachment to their mothers. Much remains to be understood about the meaning and consequences of atypical attachments. They may arise from problems parents experience in being sensitive to their child (i.e., difficulty of reading the infant's cues), from these children 's cognitive limitations, from the added stress that can accompany raising a child with special needs, or from limitations of the typical model for studying attachment when applied to these special populations. There is a tremendous need for research in this area, given its role in elucidating child factors and surrounding conditions that impinge on early attachments, as well as the developmental significance of behavioral differences in patterns of relating to important others among both atypically and typically developing children (see Vondra and Barnett, 1999).
The large majority of research on early attachments has focused on the parent-child relationship and, specifically, on the mother-infant relationship, despite the fact that young children establish close relationships with a surprising variety of people, including relatives, child care providers, and friends. Children certainly develop secure attachments to their fathers that do not depend on the security they derive from their attachments to mothers (Thompson et al., 1985). Grandmothers are also important attachment figures, and their support of the mother can facilitate secure attachment in infants (Crockenberg, 1987; Myers et al., 1987). Grandmothers are an especially important source of child care during the earliest months and years of life, as we discuss in Chapter 11.
Howes (1999) proposed the following three criteria for identification of attachment figures other than the mother: provision of physical and emotional care, continuity or consistency in the child's life, and emotional investment in the child. We do not know whether there is a specific limit to the number of people with whom very close emotional connections can be established at different ages. Regardless of their number and variety, from