Robertson Films Home Page

Background to the Separation Project

DURING THE PREVIOUS QUARTER of a century, much had been published about the general effects of separation from the mother in early childhood, mostly in the form of retrospective or follow-up studies. The few direct observational studies had been done exclusively in hospitals and other residential institutions (Burlingham and Freud, 1942, 1944; Freud, 1974; Heinicke and Westheimer, 1965; Micic, 1962; Prugh et al., 1953; James Robertson, 1953, 1953a; Schaffer and Callender, 1959; Spitz, 1945; Spitz and Wolff, 1946). Institution-based studies had the limitation that the data they provided did not permit responses to separation from the mother to be reliably differentiated from the influence of associated adverse factors such as illness, pain, cot confinement, multiple caretakers and the confusion which follows transfer from home into a strange environment. Some writers cautioned the influence of associated factors, but without being able to indicate their relative importance. For lack of means of differentiating, the literature on early separation therefore remained substantially a literature on an assortment of factors of unknown weight among which loss of the mother was only one.

This was the point at which we and John Bowlby began to see things differently. Bowlby (1960), theorizing principally on the hospital data collected by James Robertson, made what we saw as a sweeping generalization:

'that acute distress is the usual response of young children (between about six months and three to four years of age) to separation from the mother, regardless of circumstances and quality of substitute care; and, by implication, that there is no distinction between the responses of these infants at different levels of development'.

 Anna Freud (1960), criticizing Bowlby’s generalizations, involved herself in the discussion, conducted within the Psychoanalytic Study of the Child:

'Neither the Hampstead Nurseries nor hospitals and other residential homes have offered ideal conditions for the study of separation per se. We, as well as Dr Bowlby, used data collected under circumstances where the children had to adapt not only to the loss of the mother but also to the change from family to group life, a transition very difficult to achieve for any young child. Whereas the mother herself had been the undisputed possession of the child, the nurse as substitute mother had to be shared inevitably with a number of contemporaries. Also, inevitably, it is never one single nurse who substitutes for the all day and all night care of the mother'. Anna Freud stresses the lack of relevant data without which it is not possible to make statements about separation per se: ‘we need to supplement our observations, excluding group or ward conditions’.

The paediatrician and psychologist Yarrow (1961), in a definitive review of research in this area, showed that ‘maternal separation has never been studied under pure conditions’; that is, other, complicating factors were always present. Bowlby’s colleagues Heinicke and Westheimer (1965), discussing their observations on young children in a residential nursery, acknowledged that their data could not determine the influence of institutional factors, including that of multiple caretakers. They speculate that ‘if it were possible to contrast a minimal care situation with one involving highly individualized care, then one might get quite different results’.

But in Bowlby’s book Attachment and Loss (1969, Ch. 2), although there is a passing reference to the complexities of the institutional situation, there was a disappointing emphasis on the assertion that regardless of age and conditions of care the young child’s response to separation is usually the mourning sequence initiated by acute distress:

The subjects of various studies differ in many respects. For example, they differ in age, in the type of home from which they come, in the type of institution to which they go and the care they receive there, and in the length of time they are away. They differ, too, in whether they are healthy or sick. In spite of all these variations, however, and despite the different backgrounds and expectations of the observers, there is a remarkable uniformity in the findings. Once a child is over the age of six months he tends to respond to the event of separation from mother in certain typical ways.


Without citing the evidence regarding the influence of each class of variable, Bowlby asserts that ‘by far the most important variable’ is absence of the mother, and dismisses other variables as relatively unimportant. He deals thus with strange environment, previous mother—child relationship and the state of the mother as in pregnancy; and he omits to consider other variables such as quality of substitute care, multiple caretakers, age and level of maturity of the child at separation.

Moving on from our previous work, Joyce and I decided to try to clarify the subject by looking more closely at the influence of variables on the behaviour of healthy young children during a ten-day separation from the mother. In order to achieve maximum coverage we would become foster-parents to a series of young children, giving care for twenty-four hours a day and making written and filmed observations while doing so.

We made a proposal to Dr Bowlby, then director of the Tavistock Child Development Research Unit, for a project to refine the earlier studies by getting closer to separation per se. Unfortunately he was not as responsive as we would have liked.

But ultimately £1,000 of unit funds were earmarked for our use. Joyce left the Hampstead Well Baby Clinic where she had worked for the previous ten years and joined me at the Tavistock Clinic,

In a pilot study we took into our foster care Kate, aged two years five months, referred by a social worker in the maternity hospital where the mother awaited the birth of a second child. Our research experience hitherto had been of young children who had reacted with acute distress to admission to hospitals and other institutions. We were therefore apprehensive that Kate would do something similar right in our home. But she did not. For the first three or four days of the separation she was cheerful and controlled, only gradually showing signs of anxiety. There were subtleties of behaviour that were observable only because of the intimacy of her care by Joyce and never previously described in the literature.

During the month’s separation she formed a strong attachment to her substitute mother, and it was clear that in this instance the acute distress asserted by Dr Bowlby to be inevitable did not occur. We judged that this had been prevented by the quality of care during the separation. (See full case report on p. 100). We were excited by our findings, and eager to continue with the project; but unfortunately, as so often happens in research departments, other projects were considered more important and the balance of the £1,000 that had been allocated to the project was now put to other purposes. The balance having been withdrawn, we were left without funds. That was an unexpected blow which threatened to put an end to the project. It was serious, bearing in mind the difficulty of finding subjects; a second child, Jane, was already lined up for our foster care in two months’ time.

We were angered and bewildered that our project which promised to throw new light on separation behaviour was to be jettisoned, but so sure in our own minds of its value that we set about finding our own funds. Time was short. We decided to approach the Grant Foundation in New York. Adele Morrison, the general secretary, had a very positive attitude to us, partly through my work on children in hospital but more importantly because of her appreciation of the reports on the Hampstead Well Baby Clinic that Joyce had written over some years.

It happened that we were to be in the United States to lecture at Yale and Harvard, so we arranged to go first to New York for a preliminary discussion. Adele Morrison listened with keen interest to our ideas, said she would consult some colleagues and, a few days later, telephoned across the United States to say the Grant Foundation would give an advance to enable us to proceed with Jane on our return to London. In New York a few days later we picked up a cheque for £1,000 in advance and had assurance that funding for three years awaited formal appli­cation. Our satisfaction that our previous work should have merited such a positive response from the Grant Foundation, and relief that the study of Jane could go ahead, knew no bounds.

This made us independent, free of the danger of obstruction of our research. (The Grant Foundation financed this project for the next eleven years, supplementing the salary of James Robertson as a psychi­atric social worker in the National Health Service. Adele Morrison, the most knowledgeable and considerate general secretary, not only read our running accounts of observations but visited us frequently to discuss them in insightful detail — visits usually scheduled for a morning but always lasting all day. This was the most rewarding relationship with a foundation that any researchers could wish for — funding and deep interest.)

During these years we studied and filmed four young children taken into our foster care and one child filmed in a residential nursery generously funded by a personal award from Kodak. This became the series Young Children in Brief Separation.