ABSTRACT
With the advent of the 1990s, attachment theory can fairly be said to have come of age. No longer is it looked upon as primarily a variant of psychoanalysis and thus automatically suspect in scientific terms because unamenable to empirical disconfirmation. Its tenets have begun to inform ‘hard-nosed’ projects involving not only experimental designs but also biochemical variables. Many, for example, are following the lead of Hinde and his team (Hinde and Davies 1972; Hinde and Spencer-Booth 1971): McKinney and colleagues have utilized separation in nonhuman primates as a model for human depression (McKinney et al. 1984), and rat pups isolated at a critical stage of development have been shown to exhibit increased activity during wakefulness (Smith and Anderson 1984) and disorganized EEG sleep patterns (Hofer 1976). Furthermore, even in what might be considered the ‘softer’ areas of human studies utilizing verbal interviews, methods have become stricter and more elaborate: Parkes and Weiss’s discussion of recovery from bereavement in 1983 contains a wealth of statistical analyses which complement their sensitive reporting of qualitative material, whereas 9 years earlier their description of the first year of bereavement lacked the statistical cutting edge likely to convince a sceptic (Glick et al. 1974). In addition, Ainsworth and colleagues’ classic initial insights with children (Ainsworth and Wittig 1969; Ainsworth et al. 1971) have been subtly refined (Main and Weston 1982; Main and Solomon
1986), and extended beyond the Anglo-Saxon world (Grossmann and Grossmann 1981; Sagi et al. 1985; Miyake et al. 1985).
