The beauty of love has not found me
Its hands have not gripped me so tight
For the darkness of hate is upon me
I see day, not as day, but as night.
I yearn for the dear love to find me
With my heart and my soul and my might
For darkness has closed in upon me
I see day, not as day, but as night.
The children are playing and laughing
But I cannot find love in delight
There is an iron fence around me
I see day, not as day, but as night.1
FROM ALL THAT is written in earlier chapters it will already be clear that, in my view, the variables that influence the course that mourning takes during childhood and adolescence are similar in kind to those that influence it during adult life. They fall into three classes:
(a) the causes and circumstances of the loss, with especial reference to where and what a child is told and what opportunities are later given him to enquire about what has happened;
(b) the family relationships after the loss, with special reference to whether he remains with the surviving parent and, if so, how the patterns of relationship are changed as a result of the loss;
(c) the patterns of relationship within the family prior to the loss, with special reference to the patterns obtaining between the parents themselves and between each of them and the bereaved child.
Some of the evidence that supports this theoretical position has already been referred to and more is given in this and subsequent chapters. It derives from studies of two main types:
(i) studies that compare the experiences of a group of individuals who have developed well despite a childhood bereavement with those of a group who have failed to do so; information is usually obtained during a special research interview or a routine clinical one;
(ii) studies that describe the experiences of one or a few individual children or adolescents whose problems are thought to stem from the death of a parent; most of the information is obtained during therapy, though some comes from parents and others.
The strengths and weaknesses of these two types of study tend to be the opposite of one another.
Studies of the first type, which take the form of surveys, include fairly large samples of subjects and give useful information, mostly of a rather general kind, about the individual’s experience after the loss but are usually weak on psychopathological detail. Studies of the latter type, the therapeutic, do much to supplement this deficiency but can be gravely misleading when treated in isolation. In the case of surveys the information is commonly obtained many years after the events; whereas in the case of therapeutic studies of children and adolescents the time interval is usually much shorter. Both types of study have the drawback that they rely heavily on information from a single source, the bereaved individual himself.
Amongst all those who have surveyed different groups of individuals who have lost a parent during childhood there is now substantial agreement in regard to the enormous importance of a child’s experience after the loss. Individuals who later develop a psychiatric disorder, it is found, are far more likely than are those who do not to have received deficient parental care following the loss. Discontinuities of care, including being cared for in unloving foster-homes or institutions and of being moved from one ‘home’ to another, have been the lot of many. Alternatively, should a child have remained in his home, he is likely to have had to take a parental role prematurely instead of being cared for himself. By contrast, those who have developed well despite having lost a parent during childhood are likely to have received continuous and stable parental care during the years following their loss. Amongst findings that support these conclusions are those, already cited in the previous chapter, of Rutter (1966), of Adam (1973), and of Birtchnell (1971,2 1975). Amongst other studies that report closely similar findings is a well-designed one by Hilgard et al. (1960).
Hilgard, who had for many years been interested in the role of parent loss during the childhoods of psychiatric patients, with special reference to anniversary reactions, decided to compare the experiences to which her patients had been exposed following their loss with those of adults who had also lost a parent during childhood but were not patients. With this in view she undertook a community survey and from it identified one hundred individuals aged between 19 and 49 who had lost a parent before the age of nineteen and who were not at the time under psychiatric care. Of this initial sample, sixty-five made themselves available for structured interviews lasting one or two hours. Women outnumbered men by nearly three to one (partly because they were more numerous in the initial hundred and partly because they were more available for interview). Of the women, 29 had lost father and 19 had lost mother; of the men, 13 had lost father and 4 mother.
After interview a sub-sample was identified made up of all those who were deemed to be ‘reasonably well adjusted’ in terms of the following criteria: they were living in an intact home, their marriage appeared to be satisfactory, relationships with their children seemed adequate, and their scores on a brief test of social adaptation were confirmatory. Among 29 women who had lost father fourteen met these criteria.
The picture of family life before and after father’s death which emerged from the accounts they gave was as follows. Before the loss the parents had provided a stable home in which each had had a well-defined role. After the loss mother had kept the home intact but had usually had to work very hard to do so. Not only was support given to the family by its social network, but mother had proved capable of making the best use of it. ‘Strong’, ‘responsible’, ‘hardworking’ were the adjectives most used to describe her; ‘affectionate’ less often.
Should a parent have died after an illness, the children were likely to have been told of the outlook and prepared in advance for what lay in store. By these means, Hilgard writes, ‘a dying parent may convey to his child an acceptance of this complete separation and in so doing may help the child to accept it also’. Furthermore, after a father had died mothers were likely to have shared their grief with the children; and it seemed as though this had been of especial help to daughters. These family patterns, in which the children had been taken into the parents’ confidence, had contributed, Hilgard believes, to the striking absence of guilt about the parent’s death that characterized these individuals and that contrasted sharply with what she found in her group of psychiatric patients who also had lost a parent during childhood.
In addition to Hilgard’s sub-sample of individuals who were deemed to be ‘reasonably well adjusted’ was a complementary sub-sample of individuals who had failed to meet her criteria (a group analogous to the ‘community cases’ in George Brown’s study but not necessarily presenting with diagnosable illness). For members of the second sub-sample the behaviour of the surviving parent had been very different to what it had been for members of the well-adjusted sub-sample. In most cases the surviving parent had made strong demands on the children for emotional support; or, to put it in the terminology used in this work, the survivor had inverted the parent-child relationship by seeking to make his or her child the caregiver. This pattern was especially common among children whose fathers had died.
In the total sample of 65 persons interviewed, there were thirteen men who had lost father. In three cases mother had remarried, leaving ten families in which the son had continued to live with his widowed mother. In no less than nine of these cases, the mothers had ‘manifested an emotional dependency on their children, particularly the sons’. Some felt they had been made into substitute husbands. They had either stayed single until mother had died or else had married but later had divorced and returned to stay with mother. In one case mother had threatened to commit suicide when her son announced his plans to marry. Despite these pressures which had made satisfactory marriages extremely difficult, and possibly even because of them, a number of these sons had been very successful in their work. Some of the girls who had remained living with widowed mothers had also been put under strong emotional pressure to stay home to care for mother.
One of the women whose mother had died young described how, because her mother had died aged twenty-five, she had confidently expected that she also would die at that age. She had therefore postponed her marriage until after the fateful year; but she had none the less chosen for her wedding day the same date as her mother’s. When interviewed she was about forty-five and had been married for twenty years, apparently happily.
In reviewing her findings Hilgard expresses concern about members of the less well-adjusted sub-sample. Although living in the community and passing as mentally healthy, it was evident that for a number their lives had been restricted and their mental health impaired by the pathogenic pressures to which they had been subjected. Clearly some had suffered much more than others, and no doubt some of the sons who had postponed marriage had none the less married successfully later. Nevertheless, her study strongly supports the view that the effect that a parent’s death has on a child is powerfully influenced by the pattern of family relationships to which the child is exposed after it.
Certainly all the studies which have reported the childhood experiences of those who subsequently become psychiatric casualties point to the same conclusion. An example is a study by Arthur and Kemme (1964) of 83 children and adolescents, aged between 4½ and 17 years, who had been referred to a children’s psychiatric hospital in Ann Arbor, Michigan, with a variety of emotional and behavioural problems, all of which, having either developed or become greatly exacerbated following the death of a parent, could be regarded as attributable at least in part to the loss. Sixty were boys, of whom 40 had lost father and 20 mother; 23 were girls, of whom 14 had lost father and 9 mother.
Although the details given by Arthur and Kemme are rather sparse, it is evident that the conditions affecting these children and adolescents prior to the loss and/or surrounding it and/or after it had been extremely adverse in a high proportion of cases; and in many of them it was possible to see, at least in outline, how the conditions to which the child had been, or was still being subjected, were contributing to the problems complained of. Amongst the adverse conditions prominent in this series of cases were parents who had quarrelled or separated, and parents who had threatened to abandon the children, children who had experienced several earlier separations, and children who had been made to feel responsible for making the parent ill. After the death many of the children had been given little or no information about it; and subsequently many also had experienced extremely unstable relationships. Of the 83 parental deaths, 10 had been due to suicide, an incidence discussed briefly in the opening section of Chapter 22.
In the great majority of the cases reported psychological disturbance had been present before the death, often long before it. Nevertheless, in most of them it was evident that bereavement had increased any existing disturbances. As in the case of adults, therefore, the experience of loss is found to interact with the psychological consequences of both previous and subsequent adverse experiences to produce the particular clinical picture seen.
As might be expected, some of the commonest ways in which children and adolescents respond to the loss of a parent include becoming chronically sad or anxious, or some mixture of the two; and many develop elusive somatic symptoms. In the Michigan series over a quarter appeared sad at the time of referral, 16 of the 83 were showing intense separation anxiety and 19 were experiencing acute night-terrors. About a quarter were excessively clinging during the day and/or were insisting at night on sleeping with the surviving parent, or a sibling.
Yet, although many appeared obviously sad and anxious, many others did not. On the contrary, 29 children—about one-third—were overactive and in greater or less degree aggressive. Some engaged in unprovoked violence towards peers or adults or inexplicable destruction of property.
In many of the cases an explanation of a child’s sadness, anxiety or anger could be found without difficulty in the way he was construing the cause of his parent’s death and/or the situation in which he now found himself. Seventeen were construing the death in terms of their having been abandoned. As one boy put it: ‘My father left me and I’m very angry with him.’ Double that number, namely 40 per cent, were attributing the cause of death either to themselves or to the surviving parent. Several made plain why they did so. One boy, for example, had been warned by his mother that he would be the death of her. Another supposed that his mother had committed suicide because he had been so naughty. Most of those who were blaming the surviving parent had witnessed violent quarrels between their parents in which one had attacked the other physically.
Many of the younger children disbelieved that death was final and were expecting that they would soon be reunited with their parent, either here or ‘up in heaven’; some of the older children contemplated suicide with the explicit intention of joining the missing parent. Thirteen had either threatened suicide or attempted it.
In the accounts of individual children given in Chapters 19 and 21 some of these sequences are described in detail.
Throughout the last fifty years reports have been published in the psychoanalytic journals of the treatment of adult patients whose present difficulties have been thought to be due at least in part to the loss of a parent, from death or other cause, during the patient’s childhood. Since in all these cases the loss had occurred many years earlier, it is hardly surprising that the reports give little or no information about the conditions that had preceded or succeeded the loss. During the past two decades, however, accounts have multiplied of the treatment of adolescents and children whose loss had occurred comparatively recently; and in many of them a fair amount of detail is given both about the circumstances of the loss itself and about the patterns of family interaction that obtained before and after it. In the chapters to follow a number of these accounts are presented. Each has been rewritten to provide a continuous narrative shorn of extraneous theory, and with some comments of my own added.
Those sceptical of the scientific status of material obtained during the course of therapy should note that in hardly one of the cases to be described are the theoretical biases of the authors the same as my own. On the contrary, the majority subscribe more or less explicitly to the theoretical standpoint that has long been dominant amongst psychoanalysts which, until recently, has given scant weight to the influence of environmental factors and has explained almost all differences in personality development by reference to some phase of development in which the individual is thought to be fixated. When applied to differential outcome following loss this viewpoint results in the widespread assertions:
There is an extensive literature based on these premises, some of which is referred to in Chapters 1, 2 and 12 (and see also Bowlby 1960b). Readers interested to probe further are referred to a review of the literature by Miller (1971) and another most comprehensive one by Furman in the final chapter of her book (Furman 1974, especially pp. 267–93).
I believe that the evidence at present available does not support the traditional theories. A principal difficulty with some of them is that, were they to be correct, we should expect the development of every child or adolescent who lost a parent to be impaired, which we know is not the case. It is of much significance, moreover, that the closer in time to the loss that a patient, adolescent or child, has been studied and the larger the number of cases that a clinician has seen the more likely is he not only to describe environmental factors but to implicate them when explaining outcome. Among the many who now lay emphasis on environmental factors, especially the influence of the surviving parent, are the clinicians R. A. Furman (1964), E. Furman (1974), Kliman (1965), Becker and Margolin (1967), and Anthony (1973), and also the social scientists Gorer (1965), Glick et al. (1974), and Palgi (1973). The position of other clinicians seems inconsistent with the evidence they present. An example is Wolfenstein (1966, 1969) who, despite adhering strongly to traditional theory, reports evidence that seems equally strongly to implicate family relationships.3 Nagera (1970) avoids taking sides by embracing both viewpoints impartially. Thus, in a discussion of the origin of children’s beliefs that a dead father will return, he writes: ‘In some cases this happens under the direct influence of mothers who hide the truth from the child to spare it pain; in other cases phantasies of an identical nature are the child’s spontaneous production’ (italics original).
In the chapters to follow I shall repeatedly be drawing attention to the role of environmental variables, both those that the clinicians concerned refer to as having, in their opinion, been of consequence and also others that a reading of the case report has suggested to me may have been operative as well. The viewpoint adopted is, of course, consistent with the theory of developmental pathways outlined in the final chapter of Volume II and adopted throughout this work.
In judging the validity of the accounts that follow it should therefore be borne in mind that, in so far as their authors present data of a kind that support my views, it is not because of their theoretical expectations. On the contrary, I believe the reason they have done so is because in the course of their clinical work, and sometimes in spite of their theoretical bias, they have been impressed by the significance for an understanding of the children’s problems of the events they describe.
1 By a girl of eleven whose parents were abroad for some years.
2 In this study Birtchnell showed that, when a sample of early bereaved psychiatric patients was compared with a sample of general population controls, they showed an over-representation of older siblings who were of the same sex as the parent lost and who also had younger siblings to be cared for.
3 For example, in a long paper Wolfenstein (1969) describes the case of Mary who came for treatment at the age of nineteen because of being depressed, having feelings of derealization and that the best solution for everyone would be for her to kill herself. Her father had died five years earlier. Mary’s relationship to her mother had never been happy; and the mother is described as having ‘depressive tendencies’ and as prone to ‘punish the child by long silences’ (p. 444). In her comments Wolfenstein expresses the belief that Mary’s problems with her mother, together with two brief hospitalizations before the age of four, had already shaken Mary’s trust in her mother before father’s death. In spite of this, however, she concludes her paper with the generalization that ‘mourning, as a painful but adaptive process of gradually decathecting the lost object, is not an available device until after adolescence has been passed through’ (p. 457). In Chapter 21 of this volume an account is given of another adolescent girl treated by Wolfenstein in which environmental factors also appeared to have played a large part in determining outcome.