In our time, the view that parent-child relationships play a central role in a childâs psychological development has been widely accepted. Periodically there are valiant efforts to dislodge this belief from its prominent position â such as the Nurture Assumption (Harris, 1998) and Altering Fate (Lewis, 1997) â but they appear to have had little effect on the prevailing ideology. These convictions are so strong that it is hard to believe that, until recently, it has proved difficult to amass convincing empirical evidence to document the importance of parentâchild relationships in development. Attachment theory and the research that it generated played an important role in producing this evidence. Clearly the human infant is not equipped to survive without adult caregivers who provide food, warmth, and protection from illness and injury. Yet our intuitive concepts go beyond the confines of physical care and include the notion that individual differences in later functioning â indeed, the core of personality â are shaped by the experiences we have with a small number of early caregivers. The emergence of attachment theory, as articulated by John Bowlby, and the discovery by Mary Ainsworth of a way to assess individual differences in attachment behaviour patterns, laid the groundwork for intensive, ongoing, fruitful attempts to examine the psychological effects of early relationships. In this chapter, we shall trace the origins of these ideas and introduce the rest of the volume. Bretherton (1992), Grossmann (1995) and Holmes (1993) provide excellent detailed historical accounts of attachment theory and research.
Empirical roots
One logical way to understand the effects of early care and relationships is to examine what happens in their absence. Thus some of the first attempts to study the psychological effects of early care focused on the development of children raised without the benefit of a consistent caregiver, namely those reared in orphanages. In the 1940s when orphanages were common in North America, there were many such studies and they indicated not only that these children were developmentally delayed, but also that their social and emotional behaviour was unusual. They did not seem to form close relationships, and were instead described as inappropriately friendly towards everyone, including strangers (e.g. Bender and Yarnell, 1941; Goldfarb, 1943a; Skodak and Skeels, 1949; Spitz, 1945, 1946). When they were later adopted or reared in foster homes, improvements were noted in many domains, but some cognitive and affective defects persisted (e.g. Goldfarb, 1943a, 1943b, 1945).
These data were generally interpreted as the effects of âmaternal deprivationâ (Bowlby, 1951; Goldfarb, 1955; Spitz, 1956), and this view was buttressed by the now classic experiment of Skodak and Skeels (Skeels, 1966; Skodak and Skeels, 1949). They found that orphanage children who were moved to an institution for mentally retarded adults were usually âadoptedâ by a particular older resident who lavished individual attention and affection on them. These children were more likely to develop normally, leave the institution and become fully contributing members of society than those who remained in the orphanage. Now we are more aware that the privations of institutional care cannot be attributed to maternal deprivation alone. They include the absence of fathers, siblings and a family context. Nevertheless, at the time, these data played a major role in formulations concerning the motherâchild relationship. They were also a catalyst for the demise of institutional care for young children and the increasing use of foster care placements as an alternative. Ironically, when current forms of foster-care break down and children experience repeated placements, the resulting psychological effects are highly consonant with earlier descriptions of institutionalized children (Karen, 1994, chapter 1).
Other attempts to understand the role of early care in development were being made through naturalistic observations of animal behaviour as well as laboratory studies where rearing conditions could be experimentally manipulated. For example, ethologists documented the phenomenon of imprinting in precocial birds (those in which infants are mobile at birth). The term âimprintingâ was introduced by Lorenz (1935) in his studies of behaviour in geese, and there was subsequently much debate as to how imprinting should be defined (e.g. Bateson, 1966; Sluckin, 1965). The general notion is that during a âcriticalâ or âsensitiveâ period shortly after birth, exposure to a specific figure sets in motion a series of behavioural processes that result in later preferential behaviour (e.g. following, mating) towards that figure. Under normal rearing conditions, this figure is a biologically appropriate one (the mother or another species member), but experimental manipulations can lead to striking examples of imprinting on anomalous figures â for example, the famous photograph of Konrad Lorenz being followed by a line of goslings. Related behavioural phenomena were also observed in maternal behaviour. Among sheep and goats (Hersher et al., 1958), if the young are removed from the mother shortly after birth for a prolonged period, she will not accept them back into her care. Thus the initial experiences of the mother sheep or goat influence caregiving behaviour. Data such as these suggest that early contact between mother and infant has an important biological function which normally ensures or enhances caregiving and development.
In his Wisconsin Primate Laboratory, Harry Harlow and his colleagues began their studies of rhesus monkeys raised with various types of âmother surrogates.â These studies originated in a very practical problem, namely how to rear monkeys under laboratory conditions. Wire frames covered with terry cloth to which monkeys could cling for comfort, and attached bottles containing food allowed the monkeys to grow and develop in a fashion that initially appeared normal, particularly when some exposure to peers was provided (Harlow and Harlow, 1962). Later studies (Harlow and Harlow, 1965) showed that this type of rearing resulted in abnormal social behaviours, namely inability to mingle with peers and have normal sexual relations and, if females were impregnated, grossly neglecting and/or abusive behaviour towards their infants.
Bowlby and the beginnings of attachment theory
This information on the effects of early experience in the absence of normal care was accumulating at the time when John Bowlbyâs clinical experience was drawing him towards intensive consideration of the nature of early motherâchild relationships. Bowlby entered the field of child psychiatry after working as a volunteer at a school for maladjusted children, where he was struck by the impoverished family lives of some of his young charges. These observations convinced him that family life was important to emotional development and its problems, and that appropriate treatment for maladjustment required family involvement. However, this was not the custom of the time, and Bowlbyâs ideas engendered much conflict with his analytic supervisors, Melanie Klein and Joan Riviere. He also diverged increasingly from many of his psychoanalytic colleagues (who were preoccupied with internal phantasy) in believing that childrenâs actual experiences form the basis of their notions about themselves, others and relationships. Perhaps it was his childhood experience of growing up between a slightly older brother (with whom he competed) and a younger brother of more limited talents (towards whom he felt protective) that spurred Bowlby on to challenge current doctrine and champion the needs of young children (Holmes, 1993). Perhaps it was the experience of being raised by relatively distant parents who valued individual achievements, but Bowlby was determined to produce evidence in support of his position. His first empirical paper (Bowlby, 1944) was based on case-notes from experiences at the London Child Guidance Clinic. In it he presented the histories of 44 juvenile thieves to show links between their affectionless behaviour and their childhood experiences of chaotic families marked by lack of a consistent caregiver.
After World War II, he became head of the Childrenâs Department at the Tavistock Clinic, where he again found himself at odds with staff clinicians who dismissed family interactions as relatively unimportant. Consequently, he established his own research unit in order to focus on family experiences and motherâchild separation in particular. Here he collaborated with John Robertson in collecting data on hospitalization and its effects on children. Parent visiting in hospital at the time was minimal, discouraged and tightly controlled. The feeling among hospital staff was that parental visits were âupsettingâ to children, and that these upsets made the children more difficult to manage. Robertson found the grief and pain he observed in his young subjects so upsetting that he was moved to do something to help them (Holmes, 1995). As a result, Robertson (1953) produced A Two-Year-Old Goes to Hospital, a simple documentary film of the heart-wrenching experiences of one child. This film had a powerful impact and played a major role in initiating changes in hospital practices that supported and encouraged parental involvement in childrenâs care in hospital.
Soon afterwards, Bowlby was asked by the World Health Organization (WHO) to prepare a report on the mental health of homeless children. The resulting monograph, Maternal Care and Mental Health (Bowlby, 1951) was published in a popular edition, Child Care and the Growth of Maternal Love (Bowlby, 1953), and became an instant best-seller. Its major conclusion, based on the available empirical evidence, was that in order to develop normally, it is necessary for the infant/young child to experience a âwarm, intimate, and continuous relationship with his mother (or permanent mother substitute)â. However, Bowlby felt.that this statement required a theoretical explanation. His traditional theoretical orientation, namely psychoanalysis, was problematic, for despite his public success, he was a highly controversial figure and much criticized by psychoanalysts. Soon he found himself seeking out friendlier colleagues in other disciplines, especially ethology, and engaging them in ongoing seminars. His efforts at theoretical elucidation, culminating in the major three-volume work, Attachment and Loss (Bowlby, 1969, 1973, 1980), reflect these efforts to marry ethology, psychoanalysis and developmental research.
Enter Ainsworth
Mary Salter Ainsworth arrived at the Tavistock Clinic in 1950 when Bowlbyâs work on the WHO monograph had already started. She had completed her doctorate 10 years earlier at the University of Toronto as a student of William Blatz, with a thesis entitled âAn Evaluation of Adjustment on the Concept of Securityâ (Salter, 1940). Blatzâs theory of security was anchored in the belief that the experience of security within the family forms the basis from which individuals branch out as they mature and develop new skills and interests. In fact, he used the term âsecure baseâ (now so fundamental to attachment theory) to describe the role that this early family experience plays in later exploration.
While she was at the Tavistock Clinic, one of Ainsworthâs projects was to work on data from the hospitalization study. In 1953, she left for Uganda where her husband had a research position, intending to pursue the theme of motherâchild separations. She was thoroughly familiar with Bowlbyâs developing thoughts on attachment, but was not convinced that his ideas about ethology were useful to her own work on the development of infant-mother relationships (Ainsworth and Marvin, 1995). Her plan was to conduct a study among the Ganda, who traditionally separated infants from their mothers as part of the weaning process. When she found that this practice was no longer common, she reoriented her work to study the development of infant-mother attachment, and it was the resulting observations that shifted her thinking to be more consonant with that of Bowlby. This work was undertaken before Bowlby actually published the seminal papers in which he laid out attachment theory (Bowlby, 1958, 1959, 1960). When he wrote the first volume of the âAttachment Trilogyâ (Bowlby, 1969), it was one of only two existing empirical studies that informed his analysis of attachment in humans. The other, by Schaffer and Emerson (1964), was based on conducting interviews with 60 mothers of infants every 2 weeks over the first 12 months of the infantâs life and then again when the infant was 18 months old. These interviews focused on the types of infant behaviours considered to be important for attachment (separation protest, clinging, following, reaching to be picked up), and the individuals towards whom infants directed these behaviours. Ainsworth was not the first to make direct observations of infantâmother interactions, but the Ganda study and the work in Baltimore that followed it were the first studies to be framed by the concepts most relevant to attachment.
When Ainsworth returned to No...