From the
viewpoint
of this work two aspects of the case command attention.
Bowlby - Separation
At a visit six months later, during the summer holidays, the family was found in much better shape.
Mother's ulcers had healed and she had renewed contact with her married son.
Arthur was helping mother to redecorate the flat; Susan was on holiday with relatives.
When school began again both children attended more or less regularly.
This and other cases illustrate how, as soon as family inter-
-276-
viewing is adopted as a regular practice, the family origins of many intractable childhood problems are brought to light; whereas as long as each member of the family is seen only separately, interaction patterns of the greatest pathogenic significance can remain hidden. Inappropriate clinical techniques, together with strongly held theory that gives no place to the effects of family pathology, go far to explain why, with only few exceptions, those practising child psychiatry and psychoanalysis have been so very slow to recognize that a majority of children who are referred for psychiatric problems have been, and often are still being, subjected to strong pathogenic influences within their families.
Recognition of the crucial role that parents' threats play in many cases of school refusal makes it possible to read many a published case report in a new light. In some of these, for example those of E. Klein ( 1945), children are described who have a parent who is threatening that he or she will leave home or else that the child's bad behaviour will lead the parent to become ill
211
or die; despite the evidence presented, however, when the psychopathology of the children's condition is discussed, the threats are given little or no weight. In other reports cases are discussed in which it might be thought that by far the most likely explanation of why a child is afraid that harm will befall his mother is that he has heard her threaten to desert the family or to commit suicide. Yet it is clear that that possibility was never considered by the clinician, even when a child was giving the most explicit hints. As an example, one of our authors gives an interesting account of a boy of ten who told him, 'very confidentially', that one reason for his occasional reluctance to go to school was his dislike of leaving his mother alone as it was 'just possible that she might run away' and he might not find her when he returned. Yet the possibility that the boy had heard his mother make such a threat seems never to have crossed the author's mind. Another author tells of a boy who, on hearing music that reminded him of the funeral of a neighbour who had committed suicide while her child was at school, suddenly felt 'funny' and very sad, and had an irresistible urge to see his mother. The writer, after confidently explaining the fear in terms of the wish-fulfilment theory, adds, almost as an afterthought: 'There was a rather strong probability that Peter might have sensed the depressive mood of his mother and that his sudden phobia was also a sort of realistic protection of her. ' Let a spade be called a
-277-
spade: it seems more than likely that Peter had heard his mother threaten suicide.
So far in our consideration of cases falling into pattern B both the evidence presented and the argument have strongly favoured the view that refusal to go to school in such cases is a response mainly to events at home. Does this mean, then, that the wish-fulfilment theory is totally discarded? Or is it possible that the theory may have some application, even if only a limited one?
Those who support the wish-fulfilment theory very naturally point to evidence that many school-refusing children do in fact entertain hostile wishes towards a parent. And it can be agreed that, in so far as this is so, there are valid reasons for expecting the child's anxiety about his parent's safety to be increased. In some cases, therefore, the wish-fulfilment theory may apply as a partial explanation. Nevertheless, even in those cases, it is necessary to probe further, since children do not become hostile to parents for no reason.
In cases where a child is anxious about his parent's safety, not only are those who adopt the wish-fulfilment theory apt to neglect the part played by mother's threats, but they are apt also to overlook the immense frustration and provocation to which school-refusing children are often subjected. For any child to be required day after day to stay at home to keep his mother company or to make sure that she does not desert or commit suicide is the greatest of strains; and almost inevitably angry feelings are engendered. That point is made repeatedly by Johnson. In one of her papers ( Johnsonet al. 1941) she describes the treatment of a nine-year- old boy and his mother. During treatment Jack expressed much rage against his mother because of her demands upon him and because of her resentment when he strove to be independent. Almost simultaneously, during her own treatment, mother came to recognize that Jack's rages against her were an exact replica of how she herself had always reacted to the insistent demands made upon her by her own mother, who had also begrudged her doing anything on her own.
212
In conclusion, therefore, it can be said that, whenever a school-refusing child expresses anxiety about the continuing presence or safety of a parent, it is likely to be a fairly straightforward response to events occurring in his family; and that, in so far as an increased degree of anxiety stems from fear that unconscious hostile wishes may be realized, these unconscious
-278-
wishes are themselves likely to arise in response to events within the family. For these reasons, events within the family have first claim to a clinician's attention.
The remaining two patterns of family interaction probably occur less frequently than patterns A and B and can be dealt with more briefly.
Family Interaction of Pattern C
In families of pattern C a child is afraid of leaving home for fear of what might happen to himself were he to do so. Here again threats by parents, either overt or covert, usually provide the explanation.
Wolfenstein ( 1955) gives a vivid account of a case in which threats to get rid of the child were overt and, in her view, accounted for his symptoms.
Tommy, aged six years, refused to stay in nursery school or to be separated from his mother in any other way. About the time of his birth mother had lost both her parents by death, and a few months later her husband deserted. Thenceforward mother and child had lived an isolated life together. Throughout, mother was in two minds whether to keep Tommy with her or to place him in a foster home: 'While she thought constantly of getting rid of Tommy, she also clung to him desperately. He was, as she said, all she had, her whole life. ' Mother's relationship to her own mother had evidently been an extremely disturbed one; internal evidence suggests that she herself may also have been subjected to threats of being abandoned.
Mother's threats to abandon Tommy were no secret: ' Tommy not only overheard his mother discuss with neighbours the possibility of placing him, he also was repeatedly threatened with this when he misbehaved. ' Tommy's response was one of intense anxiety combined with overactive provocative behaviour and hectic laughter. During therapy he was deeply concerned that he might be sent away and often played a game in which he abandoned the therapist. To his teachers he was sometimes violent, and he shouted at them to 'Get out of here! ' In both these regards his behaviour seems clearly to have been modelled on that of his mother towards himself. Wolfenstein is in no doubt that 'the central and overpowering anxiety' in Tommy's life 'was the well-justified fear of being abandoned by his mother'. His refusal to go to school was thus a simple and intelligible response.
-279-
Robert S. Weiss (personal communication), who is studying mothers who are struggling to bring up children without a partner to help them, reports that a large proportion of them admit that, at times when they are more than usually anxious or depressed, they entertain ideas of getting rid of their children. This being so, it seems not unlikely that, in fits of desperation,
213
many of them express these ideas within earshot of their children and thereby engender deep anxieties. Unless she has very great confidence in an interviewer, however, a mother is most unlikely to admit to this.
There is in fact reason to suspect that, as in cases of pattern B, there are many children who are being subjected to threats the existence of which is kept a closely guarded secret from all those who may be called upon to help. An example, in which the secret was divulged by the child when drugged, is given by Tyerman ( 1968):
Eric was thirteen, a conscientious pupil at the technical school, and popular with both teachers and classmates. He went to church regularly with his parents and was a welcome member of the youth club. Then suddenly he refused to go to school, saying he was frightened that on the way his heart would stop beating and he would die. . . . He had read in the newspapers, he said, of people dropping dead in the street, and he was frightened that this was going to happen to him. He was eating and sleeping normally, his mother reported; but nothing seemed to interest him, and he was very preoccupied with thoughts of death. . . . His parents seemed to love each other and to love him. It appeared a happy home, and no source of tension could be discovered. There was no sign of hostility towards Eric in the school or at home, and his behaviour remained a mystery. He was not improved by taking phenobarbitone, or by talking to the psychiatrist or myself; and so an abreaction with sodium pentothal was carried out by the consultant psychiatrist.
During the abreaction Eric described a distressing event which had occurred about a week before he complained of this fear of dying. Apparently his father had accused him of stealing money out of his pockets. When Eric denied it, his father said he was going to punish him -- not for stealing, but for lying. Eric told the psychiatrist that he had not taken the money, but that he had later confessed to having done so in order to escape being beaten. When he had made his confes-
-280-
sion -- which was, in fact, his only lie -- his father said he must be punished. He drew up a document which said that he and his wife irrevocably gave up all rights to Eric, and that they wished the children's officer to take him into one of the local authority's homes. They then put Eric into the car and drove to see the children's officer. It was lunch-time and his office was closed. The boy was thereupon taken backwards and forwards from office to car until he was in tears and almost hysterical. His father then told him that as he seemed to be suitably sorry he could stay at home.
The parents did not accept invitations to come for further interviews and the boy's story remained uncorroborated. Nevertheless, those with experience in the field are likely to think the boy's story to be true, at least in substance.
Tyerman remarks that neither the parents nor the boy had mentioned the incident in earlier interviews, presumably because the parents were ashamed of their actions and the boy was afraid to tell. If we are right in thinking the boy's story true, the case illustrates yet again how very easy it is even for experienced clinicians to be misled into supposing that a child's fears have no basis in reality. It calls attention also to a main reason why clinicians have resorted so readily to theories that invoke unconscious wishes, phantasy, and projection and have been
214
correspondingly so slow to recognize the role of situational factors, either of the present or of the past.
Family Interaction of Pattern D
In families showing this pattern mother, or more rarely father, fears that something dreadful will happen to the child and so keeps him at home. In many such cases the parent's fear has been much exacerbated by the child's having been ill, occasionally seriously but more often only slightly.
Explanations of why a parent should have fears for his child again fall into two types. One, traditionally adopted by psychoanalysts, is the wish-fulfilment theory, namely that what a parent fears is that his own unconscious hostile wishes towards the child may come true. The other is that a parent is unusually apprehensive of danger befalling his child because he is reminded of some tragedy that happened in the past.
As we saw when considering the converse case of why a child should fear that harm will befall his parents, the two theories are not incompatible. In any one case either or both may apply.
-281-
Many cases of pattern D are on record in which the parent's anxiety stems from some past event. For example, Eisenberg ( 1958) describes a father whose anxiety about his son's safety was closely linked to the sudden death of his brother at the age of seventeen, for which he had felt responsible. Other examples are given by Davidson ( 1961). In one, the case of a girl of eleven, it emerged after ten months of treatment that mother's sister had died at the age of eleven. The girl herself offered this as the explanation of why she thought her maternal grandmother had suddenly become so fussy and over-protective of her. Talbot ( 1957) refers to parents still deeply preoccupied with deaths in the family that had occurred years earlier. Almost everyone practising family psychiatry who is alive to the issue will have met with similar cases.
Yet there are also cases in which the wish-fulfilment theory is certainly applicable. An example from my own experience is the intense apprehension felt by Mrs Q that Stephen might die, which was found to be a direct reaction to her own impulse to throw her baby out of the window, an impulse she had been both wholly conscious of and horrified at. What Mrs Qhad not been aware of was that her hostility to Stephen arose, in all probability, from her having redirected (displaced) towards her infant angry feelings engendered in her initially by the way her own mother behaved.
Parents of School-refusing Children: Results of Psychiatric Examination
In view of all that has been said it will come as no surprise to find, when a sample of parents of school-refusing children is examined psychiatrically, that the incidence of psychiatric disturbance is high and that, with the exception of the least severe cases, marital disharmony is universal.
Of fifty mothers studied by Hersov ( 1960b) eight had had previous psychiatric treatment (five for depressive and three for hysterical conditions) and a further seventeen were found to
215
suffer from anxiety and depression of marked degree. Of the series of thirty mothers studied by Davidson ( 1961) twelve showed symptoms of depression, including two who had been hospitalized. In a series of eighteen cases of children showing marked anxiety over separation, Britton ( 1969) reports that ten mothers had been under psychiatric treatment and another six exhibited psychiatric symptoms.
The incidence of disturbance among fathers is less pronounced though by no means negligible. Of the fifty fathers
-282-
studied by Hersov ( 1960b) eight showed psychiatric symptoms: two had had severe depressions with suicidal attempts, two others had suffered less severe depression, and another four suffered from anxiety symptoms. Davidson ( 1961) reports that eleven out of thirty fathers suffered from neurotic symptoms.
In his valuable review of the literature Malmquist ( 1965) gives much evidence of a similar kind. He is insistent that the problem is one that involves the whole family and he protests at the tendency to give too little attention to the role of father.
This completes our review of what is known of the families of children who refuse to leave home to go to school. When cases are considered in the light of the four patterns of family interaction described, it is seen, first, that, once the facts are known and the family pattern is identified, a child's behaviour is usually readily intelligible in terms of the situation he finds himself in; and, second, that many of the judgements hitherto made about such children by clinicians -- that they have been spoiled by over-indulgence, that they are afraid to grow up, that they are importunately greedy, that they wish to remain a baby tied to mother for ever, that they are fixated and regressed -- are as mistaken as they are unjust.
Two classical cases of childhood phobia: a reappraisal
In the light of our review of the family patterns that lie behind almost every case of school phobia, it is of interest to look afresh at two cases of childhood phobia that, reported during the first quarter of this century, have shaped all later theorizing. In the tradition of psychoanalysis, the classical case is that of the five-year-old Little Hans, described by Freud ( 1909). In the tradition of learning theory, a classical case is that of Peter, aged two years and ten months, described by Mary Cover Jones ( 1924b), a student of Watson.
In view of the key role that anxious attachment is held in this work to have played in all the cases of childhood phobia so far considered, is there evidence, we may ask, of its having played a part also in either of these famous cases? In what follows it is argued that in both cases there is clear presumptive evidence that it did and that, because theoretical expectations led each researcher to attend to other aspects of the case, those aspects on which weight is placed here were either overlooked or relegated to a subordinate position.
In both children the presenting symptom was an animal -283-
216
phobia. The pattern of family interaction present in the first case is likely to have been pattern B, and that in the second case pattern C.
The Case of Little Hans
A key paper in the development of psychoanalytic theorizing is Freud's study of a horse phobia in a five-year-old boy. The theory that Freud advances in that paper ( 1909) is that Little Hans's fear of being bitten by a horse had resulted from the repression and subsequent projection of his aggressive impulses, comprising hostility directed towards his father and sadism directed towards his mother. Later, he concluded: 'The motive force of the repression was fear of castration' ( 1926a, SE 20: 108). Although the origin of the hostility, oedipal or preoedipal, may have been debated by other analysts, the outline of the theory has persisted and remains the basis for all later psychoanalytic theorizing about phobias. What evidence, we may now ask, is there that anxiety about the availability of attachment figures was playing a larger part in Hans's condition than Freud realized? When the case report is read afresh in the light of our discussion of school refusal, it seems probable that anxious attachment was indeed contributing a great deal to Little Hans's problem. Most of his anxiety, it is suggested, arose from threats by his mother to desert the family. This view is advanced on two grounds:
--the sequence in which symptoms developed and statements made by Little Hans himself ( SE 10: 22-4)
--evidence in father's account that mother was in the habit of using threats of an alarming kind to discipline the boy and that those included threats to abandon him ( SE 10: 44-5).
Although the title of the paper is the 'Analysis of a Phobia in a Five-year-old Boy', Freud himself saw the child only once and the 'analysis' was conducted by Hans father. The published paper comprises father's stenographic protocol, with a running commentary and a long concluding discussion by Freud.
The parents themselves had for some years been supporters of Freud, were in fact among his first (Jones 1955), and Freud had treated mother for a neurosis before her marriage. There was a younger sister, Hanna, born three and a half years after Hans and of whom he was jealous.
Hans was four and three-quarters when father became worried
-284-
about him and consulted Freud. The problem as presented was Hans's fear that a horse would bite him in the street. Father recounted how a few days earlier Hans had been out to Scho? nbrunn with his mother, which he usually enjoyed. On this occasion, however, he had not wished to go, had cried, and had been frightened in the street when going there. On the return journey 'he said to his mother, after much internal struggling: "I was afraid a horse would bite me. "' That evening before bedtime he had remarked apprehensively: 'The horse'll come into the room. '
The symptoms, as might be expected, had not come out of the blue. According to father's record, Hans had been upset throughout the preceding week. It had begun when Hans had woken up one morning in tears. Asked why he was crying he had said to his mother: 'When I was asleep I thought you were gone and I had no Mummy to coax with. ' ( Coax was Hans expression for cuddle. ) Some days later his nursemaid had taken him to the local park, as usual. But in the street he had begun to cry and asked to be taken home, saying that he wanted to 'coax' with his mother. When later that day he had been asked why he had refused to go any further, he would not say. During that evening he had again become very frightened and cried,
217
and he had demanded to stay with his mother. The next day, his mother, eager to find out what was wrong, had taken him on the visit to Scho? nbrunn, when the horse phobia was first noticed.
Looking back further into the history, we find that the week preceding the onset of the phobia had not been the first time that Hans had expressed fear that his mother might disappear. Six months earlier, during the summer holiday, he had made remarks such as 'Suppose I was to have no Mummy' or 'Suppose you were to go away'. Looking further back still, Hans's father recalled that, when Hanna was born, Hans, aged three and a half, had been kept away from his mother. In father's opinion, Hans's 'present anxiety, which prevents him leaving the neighbourhood of the house, is in reality the longing for [his mother] which he felt then'. Freud endorses that opinion and describes Hans's 'enormously intensified affection' for his mother as 'the fundamental phenomenon in his condition' ( SE 10: 24-5; also 96 and 114 ).
Thus, both the sequence of events leading up to the phobia and Hans's own statements make it clear that, distinct from and preceding any fear of horses, Hans was afraid that his mother
-285-
might go away and leave him. Since, in the light of present knowledge, the expression of such fear should alert to the possibility that mother might have uttered threats, explicit or implicit, to leave the family, it is of interest to ask whether there is any evidence of her having done so.
Early in the record it becomes apparent that mother is inclined to use rather alarming threats. For example, when Hans is only three she is described as having threatened him that, if he touched his penis, she would send for the doctor to cut it off ( SE 10: 7-8). And we know too that, over a year later, at the time the phobia was first reported, mother was still trying to break him of the habit (p. 24 ). She is said to have 'warned him' not to touch his penis, though we are not told the nature of the warnings she was then uttering.
Three months later, however, and buried deep in the 'analytic' record, Hans lifts the curtain. He had come into father's bed one morning and in the course of talk had told his father: 'When you're away I'm afraid you're not coming home. ' Father expostulates: 'And have I ever threatened you that I shan't come home? ' 'Not you,' retorts Hans, 'but Mummy. Mummy's told me she won't come back. ' Father concedes the point. 'She said that', he replies, 'because you were naughty. ' 'Yes,' assents Hans ( SE 10: 44-5).
In the passage following father reflects, reasonably enough: 'His motive for at the most just venturing outside the house but not going away from it, and for turning round at the first attack of anxiety when he is half-way, is his fear of not finding his parents at home because they have gone away. ' Soon after, however, father reverts to an explanation along oedipal lines.
Even Hans's expressed fear that a horse might bite him is consonant with the view that fear of mother's departure was the principal source of his anxiety. This is shown by an incident that had occurred during the summer holiday of the previous year and to which Hans referred, by way of refutation, when father was trying to reassure him that horses do not bite. When Lizzi, a little girl who was staying in a neighbouring house, had gone away, the luggage had been taken to the station in a cart pulled by a white horse. Lizzi's father was there and had warned her: 'Don't put your finger to the white horse or it'll bite you' ( SE 10: 29). Thus, we find
218
Hans's fear of being bitten by a horse is closely linked in his mind to someone's departure. There is other evidence also that horses are identified with departures (e. g. p. 45 ).
-286-
On all these issues, it is evident that Freud was thinking along lines very different from those proposed here. Hans's insistent desire to remain with his mother is seen, not in terms of anxious attachment, but as the expression of his love for his mother, held to have been genitally sexual in character, having reached an extreme 'pitch of intensity' ( SE 10: 110-11). The dream that his mother had gone away and left him is held to have been, not an expression of Hans's fear that his mother would carry out a threat to desert the family, but an expression of his fear of the punishment due to him for his incestuous wishes ( SE 10: 118). The episode when Hans heard a neighbour warn that the white horse might bite is linked to a postulated wish that his father should go away, not to a fear lest his mother desert. Mother's displays of affection to Hans and her allowing him to come into bed with her are seen, not simply as a natural and comforting expression of motherly feeling, but as actions that might have encouraged, in a rather unfortunate way, Hans oedipal wishes ( SE 10: 28).
A tailpiece that tends to support the present hypothesis is that, subsequent to these events, Hans parents separated and later divorced ( SE 10: 148). (The fact that Hans was separated from his younger sister suggests that mother may have kept the little girl with her and left Hans with his father. )
There the matter must be left since there is no way of knowing which of the alternative constructions is nearer the truth. In the light of the evidence, both from the case itself and from other cases of childhood phobia reviewed earlier, the hypothesis advanced here seems no less plausible than the one adopted by Freud: it is not implausible to believe that the presenting symptom in the case of Little Hans can best be understood in terms of family interaction of pattern B.
The Case of Peter
In the literature on behaviour therapy the case of another young child, Peter, aged two years and ten months, who also suffered from intense fear of animals, has achieved some fame because it is the first recorded example of fear being deconditioned. Although the therapist, a student of Watson, assumes that the child had come to be afraid of animals through having been conditioned to fear them at some time unknown, explicit evidence regarding the way his mother treated him suggests that threats from his mother had probably played a principal part.
'When we began to study him', writes Mary Cover Jones
-287-
( 1924b), 'he was afraid of a white rat, and this fear extended to a rabbit, a fur coat, a feather, cotton wool etc. , but not to wooden blocks and similar toys. ' At the sight of a white rat in his crib ' Peter screamed and fell flat on his back in a paroxysm of fear', and he proved to be even more afraid of a rabbit. Since other children of the same age were not particularly afraid of these creatures the researchers decided to see whether they could help Peter to become less afraid of them.
219
A principal procedure used to 'decondition' Peter was for him to play each day with three other children chosen because of 'their entirely fearless attitude toward the rabbit'; during a part of the play period the rabbit was brought in. After about nine sessions, a second procedure was added: each time before the rabbit appeared Peter and the other children were given candy. Altogether some forty-five sessions were given, strung out over a period of nearly six months, during which there was a two-month interruption while Peter was in hospital for scarlet fever. From time to time Peter's progress was tested by presenting him with the rabbit when he was alone. At the end of the process Peter was no longer afraid of the rabbit or of the feather and he was much less afraid of the rat and the fur coat.
From the viewpoint of this work two aspects of the case command attention.
First, Peter is described as having come from a disturbed family that was living in impoverished conditions. Throughout the experiment, it seems, he was in a residential nursery 1 or else a hospital. His mother is described as 'a highly emotional individual who cannot get through an interview without a display of tears'. Peter's older sister had died and the parents were said thenceforward to have lavished 'unwise affection' on him. Discipline was 'erratic' and in her attempts to control him mother is said to have resorted to threats: the example given is 'Come in Peter, someone wants to steal you'. The pattern of family interaction suggested by this limited information is pattern C.
The second point of interest is the effect on the deconditioning process of the presence or absence of a particular student assistant of whom Peter was fond and who he insisted was his father.
____________________
1 Although it is not stated explicitly that Peter was resident in the nursery in which
deconditioning took place, a phrase (near the end of the paper) that 'he has gone home to a difficult environment' suggests that this was so.
-288-
On each of two occasions when this assistant was present Peter became decidedly less afraid, although the assistant made no overt suggestions. On this phenomenon Jones comments, 'it may be that having him there contributed to Peter's general feeling of well-being and thus indirectly affected his reactions'.
Animal phobias in childhood
There is no disposition to argue here that every case of animal phobia in childhood and later life is but the tip of an iceberg the great bulk of which comprises intense fear of losing an attachment figure. In some individuals, no doubt, an animal phobia has developed because as children they had some frightening experience in which they were attacked by an animal of that species. In other cases, seeing or hearing about such events, perhaps in dramatic circumstances and at an age when misunderstanding and fallacious over-generalization are common, may be responsible. In yet others, prolonged exposure to a parent or other adult who habitually responds with fear to a particular species of animal may play a part. Whatever the causes, Marks ( 1969) presents evidence suggesting that there are individuals who are acutely afraid of a particular species of animal but who do not suffer from any other form of emotional disturbance.
220
Yet, although cases of true and limited animal phobia may well exist, there can be little doubt that, in very many children and probably adults also who are more than usually afraid of animals, the principal source of anxiety lies in the home and not outside it. It is already suggested that the cases of Little Hans and Peter can usefully be considered in that light. Further and substantial evidence stems from the finding that, as already described, many school-refusing children number among their heterogeneous symptoms a fear of animals. Furthermore, just as any expressed fear of school sinks into oblivion once the disturbed family situation is recognized and dealt with, so does any expressed fear of animals. Because that is so and because difficulties at home are so frequently kept a secret, it is wise when confronted by a patient complaining of animal phobia always to examine carefully the pattern of interaction within the family from which the patient comes.
The wisdom of this course is well illustrated by a case of animal phobia in an adult reported by Moss ( 1960). The patient was a woman of forty-five who had suffered since childhood from an intense fear of dogs. After seeing a film ( The Three Facesof Eve
-289-
of Eve) in which a woman is treated for phobia by means of hypnosis, she sought hypnotic treatment for herself.
During the course of treatment the patient recalled a tragic event that had occurred when she was aged four. It appeared that she had been playing in the backyard of her home with her younger sister when the family dog, Rover, knocked the little sister down. A splinter entered the child's cheek, the wound went septic and a few days later the child died. The patient recalled how her mother had accused her of having knocked her sister down and had openly blamed her for the death, and also how thenceforward she had deeply disliked Rover and had become afraid of dogs of every kind. A few years later, after another sister had been born, she recalled that she had been much afraid lest a puppy attack that sister also.
After the patient had recalled how her mother had blamed her for her sister's death, much in her life seemed to her to fit into place. For example, the episode seemed to explain, at least in part, why she had always felt misunderstood by her mother, why she had suffered from a chronic sense of guilt and a compulsive desire to please, and why her relationship to her mother had been so deeply ambivalent.
When an event that a patient recalls has occurred many years previously it is extremely difficult to be sure how valid the recalled details may be. In this case it was possible to obtain limited corroboration of the patient's story. An elder brother confirmed the presence of Rover and also that at the time of the fatal accident his two little sisters had been alone, because he and his brother, who seem to have been left in charge, had gone off to watch a fire. The patient's youngest sister recalled how in later years the patient had anxiously protected her from the approach of any and every dog. There was, however, no corroboration that mother had blamed the patient for the accident, and mother herself, who was still alive, denied having done so.
Experience in family psychiatry shows, nevertheless, that, when a young child dies, it is by no means uncommon for a parent, distraught by what has happened and perhaps feeling guilty over failure to have taken some precaution, impetuously to attribute blame to an older child. In some families the older child then becomes a scapegoat; in others the parent, after
221
recovering from the shock of acute grief, may forget, and then deny, having ever made the accusation. But in either case the accusation cuts deep, even when memory of it is repressed.
-290-
That is what seems likely to have occurred in the case described. If that were so, there would be little wonder that the accused child had come to hate and fear the animal that she believed responsible for her disgrace. Nor would there be wonder that she should have felt thenceforward that her mother, and therefore all others to whom she might look for comfort and support, would disown her and treat her with nothing but contempt.
Enough has been said perhaps to show that the theory of anxious attachment outlined in earlier chapters can illuminate many a case in which a child is intensely and persistently afraid of some situation in circumstances that are perplexing to all around him and perhaps also to the child himself. In the next chapter the problem of agoraphobia in adults is considered in the light of the same theory.
-291-
Chapter 19
Anxious Attachment and 'Agoraphobia'
It follows from the nature of the facts . . . that we are obliged to pay as much attention in our case histories to the purely human and social circumstances of our patients as to the somatic data and the symptoms of the disorder. Above all, our interests will be directed towards their family circumstances . . .
SIGMUND FREUD ( 1905a)
Symptomatology and theories of 'agoraphobia'
When a psychiatrist used to dealing with children and families examines the problem of 'agoraphobia' 1 he is at once struck by its resemblance to school phobia. In both types of case the patient is alleged to be afraid of going into a place filled with other people; in both the patient is apt to be afraid of various other situations as well; in both the patient is prone to anxiety attacks, depression, and psychosomatic symptoms; in both the condition is precipitated often by an illness or death; in both the patient is found to be 'overdependent', to be the child of parents one or both of whom suffer from long-standing neurosis, and frequently also to be under the domination of an 'overprotective' mother. Finally, a significant number of agoraphobic patients were school refusers as children.
Although minor degrees of agoraphobia are probably common and, when of recent origin, probably have a high remission rate ( Marks 1971), patients who come to the attention of psychiatrists are usually those who are suffering either from a chronic condition of some severity or else from an acute attack. Often a patient is intensely anxious, apt to panic when unable
____________________
1 The condition under discussion appears in the literature under many names, including
222
anxiety hysteria, anxiety neurosis, anxiety state, and phobic anxiety-depersonalization syndrome ( Roth 1959). The name most widely adopted at present is agoraphobia ( Marks 1969). Since criteria used to select cases differ from study to study, the extent to which findings are comparable remains in doubt.
-292-
to get home quickly, and to be afraid of an extraordinarily broad range of situations (typically, crowded places, the street, travelling) or of collapsing or even dying when out on his own. From among this heterogeneous and variable collection of situations feared it is possible, none the less, to identify two that are feared in virtually every case and are also the most feared. These situations are, first, leaving familiar surroundings and, second, being alone, especially when out of the house. Since the argument advanced here turns on the fact that fear of these situations is at the heart of the syndrome, let us consider the evidence.
During the past decade there has been very active interest in the syndrome by psychiatrists in the United Kingdom. Roth and his colleagues in Newcastle upon Tyne describe two series of cases, each numbering over one hundred ( Roth 1959; 1960; Harper & Roth, 1962; Roth, Garside & Gurney 1965; Schapira, Kerr & Roth 1970). Special aspects of the condition to which they give attention are: the high incidence of traumatic precipitating events, notably actual or threatened physical illness, bereavement and illness in the family; the high incidence of depersonalization; and the close relation of the condition to states of anxiety and depression. Another programme of research into the condition, with special reference to the efficacy of different methods of treatment, is one conducted at the Maudsley Hospital, London, by Marks and Gelder (for references to their numerous papers see Marks 1969 and 1971). A third study of value is by Snaith ( 1968) who reports on forty-eight cases of phobia in adult patients, twenty-seven of whom were typically agoraphobic. Roberts ( 1964) describes results of a follow-up of thirty-eight patients, all married women.
Although none of these workers approaches the problem from a standpoint in any way similar to that adopted here, each endorses the view that a principal feature of the condition is fear of leaving home. Roth ( 1959) speaks of 'a fearful aversion to leaving familiar surroundings'; Marks ( 1969) holds that 'fear of going out is probably the most frequent symptom from which others develop'; Snaith ( 1968) finds that, in twentyseven of his forty-eight cases, the principal source of fear is leaving home and its attendant circumstances. Furthermore he reports, first, that the more anxious an agoraphobic patient becomes the more intense grows his fear of leaving home and, second, that when a patient becomes more anxious his fear of leaving home is magnified in intensity by a factor many times
-293-
greater than is his fear of anything else. These findings lead Snaith to suggest that the condition is not a true phobia and that a more appropriate label for it would be 'non-specific insecurity fear'. In keeping with Snaith's perspective is the criterion that Roberts ( 1964) laid down for inclusion in his series, namely a patient's inability to leave his house without a companion.
Not only do these workers find that fear of leaving home when unaccompanied is the principal feature of agoraphobia, but each of them reports also that most patients have been anxious individuals all their lives: some for decades have been uneasy about going out alone (
223
Marks 1969). Between 50 and 70 per cent of patients are reported to have suffered from fears and phobias during their childhood ( Roth 1960; Roberts 1964; Snaith 1968). In a recent survey of 600 cases using a questionnaire, between one-fifth and one-sixth described themselves as having been in some degree 'school phobic' ( Berg, Lipsedge & Marks, in preparation).
Again, although psychoanalysts working in the classical tradition have an approach to the problem entirely different from that of any of the workers so far cited, and different also from that adopted in this work, they report almost exactly the same findings. For example, in an early paper describing the case of a small boy, Abraham ( 1913) notes that the boy 'does not speak of fear, but of his desire to be with his mother'. This leads Abraham to conclude that the basic problem in patients suffering from agoraphobia is that their 'unconscious . . . does not permit them to be away from those on whom their libido is fixated'.
Both Deutsch ( 1929) and, in recent years, Weiss ( 1964) endorse Abraham's view. Weiss notes especially that a patient's anxiety is apt to increase the further from home he goes, which leads him to define agoraphobia as 'an anxiety reaction to abandoning a fixed point of support'.
Thus, despite great variation in the approach and outlook of these many workers, the findings they report are impressively consistent. Only when attempts are made to accommodate the findings within a theoretical framework do differences and difficulties arise.
Three Types of Theorizing
Here, as so often elsewhere, the two rival types of theory that dominate the field are psychoanalytic theory and learning
-294-
theory. In the case of agoraphobia, however, a third type of theory has also been advanced, namely Roth's psychosomatic theory which invokes both psychological and neurophysiological processes ( Roth 1960). Strikingly enough, despite all the telltale hints that a major part is played by relationships within the patient's family of origin, a fourth type of theory, namely one that invokes pathogenic patterns of family interaction as major aetiological agents, is conspicuous by its absence.
1. Psychoanalytic theories of agoraphobia come in two main variants according to whether they focus on fear of being in the street or fear of leaving home.
Freud tends to concentrate on fear of being in the street, which he sees as a displacement outward of the patient's fear of his own libido. Even though in 1926 Freud began a major revision of his views and reached the conclusion that 'the key to an understanding of anxiety' is 'missing someone who is loved and longed for' (see Chapter 2 of this volume), he never applied his new theory to agoraphobia. 1 As a result, his original hypothesis continues to be invoked by a number of psychoanalysts who still see sexual temptation, of one kind or another, as the principal situation that an agoraphobic patient fears (e. g. Katan 1951; Friedman 1959; Weiss 1964).
224
Other psychoanalysts in their theorizing take as their focus a patient's fear of leaving home and, in doing so, advance theories very similar to those their colleagues advance to account for the similar fear found in children diagnosed as suffering from school phobia. Thus Deutsch ( 1929) notes that the reason an agoraphobic patient feels compelled to remain near his mother (or other loved person) is that he entertains unconscious hostile wishes against her and so has to remain with her to ensure that his wishes are not enacted. For Weiss ( 1964) the patient's urge to remain at home is to be understood as due to a 'regression to unresolved dependency needs'. This is also the view of Fairbairn ( 1952), although in his case histories he attributes a causal role to the very insecure childhoods his patients had experienced.
In none of the psychoanalytic formulations, apart perhaps
____________________
1 In one of his last works, New Introductory Lectures ( 1933), Freud writes: 'the agoraphobic
patient is . . . afraid of feelings of temptation that are aroused in him by meeting people in the street. In his phobia he brings about a displacement and henceforward is afraid of an external situation' ( SE 22: 84).
-295-
from Fairbairn's, is there any suggestion that a patient's refusal to leave home is a response to the behaviour of one of his parents, not only behaviour that may have occurred at some time during the past but behaviour that may be occurring still in the present.
2. During the past decade or so a new approach to a theoretical understanding of phobic conditions of all sorts has been made, this time by learning theorists; and formulations that attempt to account for each of the various situations feared have been advanced. Whereas this approach may well help us to understand some of the discrete animal phobias, how much it can contribute to an understanding of agoraphobia remains in doubt. Describing the present position as he sees it Marks ( 1969), who has made a special study of agoraphobia and draws extensively on learning theory, writes as follows:
Certain phobias, especially agoraphobia, are commonly found together with multiple other symptoms such as diffuse anxiety, panic attacks, depression, depersonalization, obsessions and frigidity. Learning theory does not explain why these symptoms develop, why they occur together, nor why they are associated more often with agoraphobia than with any other type of phobia.
Furthermore, in Marks's view, 'the origin of the panics, depression and other symptoms is not indicated by learning theory' (p. 93 ).
How the panics and depressions do originate is, for Marks, the most puzzling aspect of the condition. For, in his opinion, not only is learning theory unable to account for them but no other theory can do so either (p. 93 ). Admitting the quandary, Marks leaves the matter open; but he tends to the view that anxiety attacks probably have an unknown physiological origin. Nowhere does he consider the possibility that they may originate in family situations that create psychological distress.
Having recognized frankly the difficulties in accounting for agoraphobic symptoms entirely in terms of learning theory, Marks believes nevertheless that the theory has much to offer. The
225
hypothesis he advances is based on the idea, suggested by learning theory, 'that panic attacks and depression [may] act as super-reinforcers which facilitate phobic conditioning' whenever a patient who happens to be experiencing such affects goes out of his home. This line of thought leads Marks to propose that, in the development of agoraphobia, the anxiety
-296-
attack comes first and the situations that the patient reports he fears come to be feared only later, either as a result of a secondary conditioning effect or as a result of rationalization. In that context both fear of going out of the house and fear of becoming separated from a companion, the two symptoms most characteristic of agoraphobic patients, are held to develop through a process of secondary conditioning.
In keeping with his hypothesis, Marks expresses much scepticism regarding the causal role of precipitating factors, holding that they probably act simply as 'non-specific stressors in a patient already liable to the disorder . . . or that the disorder was already present, but hidden until the stressor elicited or exacerbated it'. In support of his position he lays much emphasis on his claim that 'not a few phobias start suddenly without any obvious change in the life- situation of the patient' (p. 128 ).
Both the sequence of events that Marks postulates and the weakness of his position are illustrated in his description of the case of a woman who sought treatment at the age of thirtythree on account of depression with suicidal ideas. The account she then gave was that, ten years earlier when she was aged twenty-three, she had developed anxiety, sweating, and shaking of the legs while travelling to work by train. Subsequently she had discovered that she felt better if her husband was present and so had taken a job in the firm in which he worked. After a few months, however, she had become afraid of separation from him, had to know exactly where he was and had telephoned him frequently. If for any reason she could not contact him immediately she would panic, feel completely lost and want to scream.
The only information Marks gives regarding this patient's childhood is that 'as a small child [she] used to be frightened when her parents were out and once sent out her younger brother to find them. She had infrequent desires to scream which were hard to stifle. These disappeared in her late teens. '
Despite the uncertain validity of the retrospective data, Marks seems confident about the sequence of symptoms: 'First came the travel phobia and depersonalization, then came the discovery of relief in the presence of her husband and after this he became indispensable. Finally the patient presented for treatment of separation anxiety. ' In accounting for the symptoms Marks proposes two distinct pathologies. On the one hand is the agoraphobia and on the other is the anxiety about
-297-
separation, to which he believes the patient had been sensitized as a child. Originating independently, the two pathologies are held subsequently to have interacted.
There are several flaws in Marks's position. First, in the light of the childhood history of this patient, it is difficult to accept his confident assertion that agoraphobia came first and separation anxiety afterwards. Second, in his ready acceptance of this and other patients'
226
accounts that the initial anxiety attack came 'out of the blue', he makes no allowance for a patient's witting or unwitting suppression of information, a process we know to be extremely common and often to hide clues vital for understanding the condition. Third, to postulate two distinct psychopathologies for a pair of symptoms that habitually go together 1 is far from parsimonious. Finally, as Marks himself admits, he can give no explanation of how or why this patient (or any other) first started to experience anxiety and panic attacks.
An alternative hypothesis to account for this patient's symptoms is that, during her childhood, she had been subjected to repeated and realistic threats of being abandoned, so that even after reaching adult life she had continued to be acutely sensitive to any such danger. The case is discussed further below (p. 307 ).
As regards the part played by a patient's parents in the genesis of agoraphobia, learning theorists share with traditional psychoanalysts the same shadowy picture. Whereas neither group attributes much importance to parental behaviour, in so far as they do so both invoke the theory of spoiling. As Andrews ( 1966) points out, Wolpe ( 1958) and Lazarus ( 1960), two leading learning theorists, both regard a patient's tendencies to withdraw to and to remain at home as responses he has learnt during interaction with over-protective parents. Marks ( 1969) in his discussion of prevention implies the same process. Some years earlier Terhune ( 1949), a psychiatrist whose outlook is in many ways similar to that of present-day learning theorists, wrote confidently: 'The phobic person is one who has been over-protected, brought up "soft". '
3. The third main type of theorizing about agoraphobia,
____________________
1 Marks rests his argument on his finding that about 5 per cent of agoraphobic patients are
not helped by company and prefer to be alone while travelling ( Marks 1969: 98). in most syndromes, however, cases occur that lack one or more typical symptoms; an example is measles without a rash. Such atypical cases require special study.
-298-
initiated by Roth ( 1959; 1960), regards the condition as being truly psychosomatic. In presenting his theory, Roth lays much emphasis on the vulnerable personalities of his patients, on the precipitating role of stressful events, and on depersonalization, which he regards as a core symptom of the syndrome. The psychological factors to which he points include both situations that, acting perhaps from early childhood, are thought to have contributed to the development of an anxious dependent personality, and stressful events, such as bereavement and illness, that appear to act as precipitants. The somatic factor he postulates is a specific cerebral mechanism that, once triggered, is difficult to inactivate. After considering certain disturbances of perception and consciousness that he finds in these patients, together with symptoms that he attributes to temporal lobe dysfunction, Roth concludes that the somatic pathology is probably to be understood as arising in the mechanisms regulating awareness, which he postulates to have become chronically deranged. Although he gives little detail of how he believes a difficult childhood and stressful situations of later life interact to produce an agoraphobic syndrome, Roth's approach is not incompatible with that presented here.
We turn now to consider a fourth type of theory, namely the one that results when the problem of agoraphobia is looked at in the theoretical perspective developed in this work.
227
Throughout our further discussion it is important to bear in mind that, as all workers now agree, the central symptom of the condition under scrutiny is fear of leaving home.
Pathogenic patterns of family interaction
Whether the theory of anxious attachment, applied already to problems of school phobia, can help to solve problems of agoraphobia also must remain in doubt. For, apart from some limited and mostly rather crude observations, there are few data yet available on patterns of interaction within an agoraphobic patient's family of origin. Almost all the data yet published come either from the patient himself or from a single interview with a relative, with the many distortions and omissions that we know such clinical procedures entail. What is missing, but so necessary to have, is first-hand observations of how a patient and his parents are currently behaving towards one another. In the absence of such data all that is possible here is to draw attention to certain reasonably well-attested findings
-299-
that are at least consonant with the view that many, if not all, cases of agoraphobia can be understood as products of pathogenic patterns of family interaction. There is much evidence of a rather general kind which, although it gives little information about specific patterns of interaction, points to a high incidence of disturbance in the families from which agoraphobic patients come. Before considering specific patterns, therefore, we consider this general evidence. Most reports agree that a majority of agoraphobic patients come from homes that are intact, in the sense that there are two parents living continuously together. Yet there is also substantial evidence that, within these homes, relationships are often far from harmonious; and it is repeatedly indicated that the parents of patients are neurotic or disturbed in some other way. Taking as a criterion clear-cut neurotic breakdown in first-degree relatives, Roth ( 1959) finds an incidence of 21 per cent. Nor should it be overlooked that a minority of patients, in one study as high as 25 per cent, come from homes broken by death, divorce, or other cause. In a recent study of eighty-seven consecutive London patients by M. S. Lipsedge (unpublished), a high incidence of disturbance in the families of origin is reported. The patients ranged in age from twenty-two to sixty-four years; fourteen were men and seventy- three women. Almost all the information was obtained from the patients themselves during an initial interview, though occasionally it was supplemented by information from a general practitioner. Inadequate though this method is for obtaining the information required, it is unlikely to exaggerate the degree of disturbance in the families from which the patients came. On the basis of this information, the patients' families can be divided very roughly into three categories:
I. I intact and reasonably stable families
II. II intact families in which there was much quarrelling, violence, and alcoholism, and/or
almost complete absence of affection
III. III families broken by death or divorce, or in which a parent was chronically ill, and/or
the patient had had prolonged separations from or changes of parent figure.
The number and proportion of patients from families in each category are shown in the following table.
-300-
228
Category of No.
of % of patients
42 30 28 100
family
I
II
III Total
patients
37 26 24 87
Category I: There were thirty-seven patients who described their home life as having been happy or who gave no particularly adverse information about it. Nevertheless, two of these patients described a parent as having been agoraphobic (one father and one mother), and two others described themselves as having been 'over-protected'. Ten of the patients referred to themselves as having been notably fearful as children; of these, two had been school refusers and one agoraphobic. Thus neurotic trouble of one kind or another is reported in members of about one-third of these not overtly disturbed families.
Category II: There were twenty-six patients whose homes appear to have been intact but who described themselves as having had an extremely unhappy family life when children. Eighteen patients described their parents as having engaged in perpetual quarrels, including violence, and often made worse by alcohol. Another eight complained of having received no affection and/ or of having been rejected. In three of these twenty-six cases the patient's mother had been agoraphobic. Two of the patients had themselves been school refusers as children.
Category III: Of the remaining twenty-four patients, there were twenty-one whose family life had been disrupted by death, divorce, or desertion and/or who had experienced many changes of mother figure. Of these, ten had lost one or both parents by death before their tenth birthday (six a father, three a mother, and one both). In five cases mother had deserted and in at least one other father.
This and other cases illustrate how, as soon as family inter-
-276-
viewing is adopted as a regular practice, the family origins of many intractable childhood problems are brought to light; whereas as long as each member of the family is seen only separately, interaction patterns of the greatest pathogenic significance can remain hidden. Inappropriate clinical techniques, together with strongly held theory that gives no place to the effects of family pathology, go far to explain why, with only few exceptions, those practising child psychiatry and psychoanalysis have been so very slow to recognize that a majority of children who are referred for psychiatric problems have been, and often are still being, subjected to strong pathogenic influences within their families.
Recognition of the crucial role that parents' threats play in many cases of school refusal makes it possible to read many a published case report in a new light. In some of these, for example those of E. Klein ( 1945), children are described who have a parent who is threatening that he or she will leave home or else that the child's bad behaviour will lead the parent to become ill
211
or die; despite the evidence presented, however, when the psychopathology of the children's condition is discussed, the threats are given little or no weight. In other reports cases are discussed in which it might be thought that by far the most likely explanation of why a child is afraid that harm will befall his mother is that he has heard her threaten to desert the family or to commit suicide. Yet it is clear that that possibility was never considered by the clinician, even when a child was giving the most explicit hints. As an example, one of our authors gives an interesting account of a boy of ten who told him, 'very confidentially', that one reason for his occasional reluctance to go to school was his dislike of leaving his mother alone as it was 'just possible that she might run away' and he might not find her when he returned. Yet the possibility that the boy had heard his mother make such a threat seems never to have crossed the author's mind. Another author tells of a boy who, on hearing music that reminded him of the funeral of a neighbour who had committed suicide while her child was at school, suddenly felt 'funny' and very sad, and had an irresistible urge to see his mother. The writer, after confidently explaining the fear in terms of the wish-fulfilment theory, adds, almost as an afterthought: 'There was a rather strong probability that Peter might have sensed the depressive mood of his mother and that his sudden phobia was also a sort of realistic protection of her. ' Let a spade be called a
-277-
spade: it seems more than likely that Peter had heard his mother threaten suicide.
So far in our consideration of cases falling into pattern B both the evidence presented and the argument have strongly favoured the view that refusal to go to school in such cases is a response mainly to events at home. Does this mean, then, that the wish-fulfilment theory is totally discarded? Or is it possible that the theory may have some application, even if only a limited one?
Those who support the wish-fulfilment theory very naturally point to evidence that many school-refusing children do in fact entertain hostile wishes towards a parent. And it can be agreed that, in so far as this is so, there are valid reasons for expecting the child's anxiety about his parent's safety to be increased. In some cases, therefore, the wish-fulfilment theory may apply as a partial explanation. Nevertheless, even in those cases, it is necessary to probe further, since children do not become hostile to parents for no reason.
In cases where a child is anxious about his parent's safety, not only are those who adopt the wish-fulfilment theory apt to neglect the part played by mother's threats, but they are apt also to overlook the immense frustration and provocation to which school-refusing children are often subjected. For any child to be required day after day to stay at home to keep his mother company or to make sure that she does not desert or commit suicide is the greatest of strains; and almost inevitably angry feelings are engendered. That point is made repeatedly by Johnson. In one of her papers ( Johnsonet al. 1941) she describes the treatment of a nine-year- old boy and his mother. During treatment Jack expressed much rage against his mother because of her demands upon him and because of her resentment when he strove to be independent. Almost simultaneously, during her own treatment, mother came to recognize that Jack's rages against her were an exact replica of how she herself had always reacted to the insistent demands made upon her by her own mother, who had also begrudged her doing anything on her own.
212
In conclusion, therefore, it can be said that, whenever a school-refusing child expresses anxiety about the continuing presence or safety of a parent, it is likely to be a fairly straightforward response to events occurring in his family; and that, in so far as an increased degree of anxiety stems from fear that unconscious hostile wishes may be realized, these unconscious
-278-
wishes are themselves likely to arise in response to events within the family. For these reasons, events within the family have first claim to a clinician's attention.
The remaining two patterns of family interaction probably occur less frequently than patterns A and B and can be dealt with more briefly.
Family Interaction of Pattern C
In families of pattern C a child is afraid of leaving home for fear of what might happen to himself were he to do so. Here again threats by parents, either overt or covert, usually provide the explanation.
Wolfenstein ( 1955) gives a vivid account of a case in which threats to get rid of the child were overt and, in her view, accounted for his symptoms.
Tommy, aged six years, refused to stay in nursery school or to be separated from his mother in any other way. About the time of his birth mother had lost both her parents by death, and a few months later her husband deserted. Thenceforward mother and child had lived an isolated life together. Throughout, mother was in two minds whether to keep Tommy with her or to place him in a foster home: 'While she thought constantly of getting rid of Tommy, she also clung to him desperately. He was, as she said, all she had, her whole life. ' Mother's relationship to her own mother had evidently been an extremely disturbed one; internal evidence suggests that she herself may also have been subjected to threats of being abandoned.
Mother's threats to abandon Tommy were no secret: ' Tommy not only overheard his mother discuss with neighbours the possibility of placing him, he also was repeatedly threatened with this when he misbehaved. ' Tommy's response was one of intense anxiety combined with overactive provocative behaviour and hectic laughter. During therapy he was deeply concerned that he might be sent away and often played a game in which he abandoned the therapist. To his teachers he was sometimes violent, and he shouted at them to 'Get out of here! ' In both these regards his behaviour seems clearly to have been modelled on that of his mother towards himself. Wolfenstein is in no doubt that 'the central and overpowering anxiety' in Tommy's life 'was the well-justified fear of being abandoned by his mother'. His refusal to go to school was thus a simple and intelligible response.
-279-
Robert S. Weiss (personal communication), who is studying mothers who are struggling to bring up children without a partner to help them, reports that a large proportion of them admit that, at times when they are more than usually anxious or depressed, they entertain ideas of getting rid of their children. This being so, it seems not unlikely that, in fits of desperation,
213
many of them express these ideas within earshot of their children and thereby engender deep anxieties. Unless she has very great confidence in an interviewer, however, a mother is most unlikely to admit to this.
There is in fact reason to suspect that, as in cases of pattern B, there are many children who are being subjected to threats the existence of which is kept a closely guarded secret from all those who may be called upon to help. An example, in which the secret was divulged by the child when drugged, is given by Tyerman ( 1968):
Eric was thirteen, a conscientious pupil at the technical school, and popular with both teachers and classmates. He went to church regularly with his parents and was a welcome member of the youth club. Then suddenly he refused to go to school, saying he was frightened that on the way his heart would stop beating and he would die. . . . He had read in the newspapers, he said, of people dropping dead in the street, and he was frightened that this was going to happen to him. He was eating and sleeping normally, his mother reported; but nothing seemed to interest him, and he was very preoccupied with thoughts of death. . . . His parents seemed to love each other and to love him. It appeared a happy home, and no source of tension could be discovered. There was no sign of hostility towards Eric in the school or at home, and his behaviour remained a mystery. He was not improved by taking phenobarbitone, or by talking to the psychiatrist or myself; and so an abreaction with sodium pentothal was carried out by the consultant psychiatrist.
During the abreaction Eric described a distressing event which had occurred about a week before he complained of this fear of dying. Apparently his father had accused him of stealing money out of his pockets. When Eric denied it, his father said he was going to punish him -- not for stealing, but for lying. Eric told the psychiatrist that he had not taken the money, but that he had later confessed to having done so in order to escape being beaten. When he had made his confes-
-280-
sion -- which was, in fact, his only lie -- his father said he must be punished. He drew up a document which said that he and his wife irrevocably gave up all rights to Eric, and that they wished the children's officer to take him into one of the local authority's homes. They then put Eric into the car and drove to see the children's officer. It was lunch-time and his office was closed. The boy was thereupon taken backwards and forwards from office to car until he was in tears and almost hysterical. His father then told him that as he seemed to be suitably sorry he could stay at home.
The parents did not accept invitations to come for further interviews and the boy's story remained uncorroborated. Nevertheless, those with experience in the field are likely to think the boy's story to be true, at least in substance.
Tyerman remarks that neither the parents nor the boy had mentioned the incident in earlier interviews, presumably because the parents were ashamed of their actions and the boy was afraid to tell. If we are right in thinking the boy's story true, the case illustrates yet again how very easy it is even for experienced clinicians to be misled into supposing that a child's fears have no basis in reality. It calls attention also to a main reason why clinicians have resorted so readily to theories that invoke unconscious wishes, phantasy, and projection and have been
214
correspondingly so slow to recognize the role of situational factors, either of the present or of the past.
Family Interaction of Pattern D
In families showing this pattern mother, or more rarely father, fears that something dreadful will happen to the child and so keeps him at home. In many such cases the parent's fear has been much exacerbated by the child's having been ill, occasionally seriously but more often only slightly.
Explanations of why a parent should have fears for his child again fall into two types. One, traditionally adopted by psychoanalysts, is the wish-fulfilment theory, namely that what a parent fears is that his own unconscious hostile wishes towards the child may come true. The other is that a parent is unusually apprehensive of danger befalling his child because he is reminded of some tragedy that happened in the past.
As we saw when considering the converse case of why a child should fear that harm will befall his parents, the two theories are not incompatible. In any one case either or both may apply.
-281-
Many cases of pattern D are on record in which the parent's anxiety stems from some past event. For example, Eisenberg ( 1958) describes a father whose anxiety about his son's safety was closely linked to the sudden death of his brother at the age of seventeen, for which he had felt responsible. Other examples are given by Davidson ( 1961). In one, the case of a girl of eleven, it emerged after ten months of treatment that mother's sister had died at the age of eleven. The girl herself offered this as the explanation of why she thought her maternal grandmother had suddenly become so fussy and over-protective of her. Talbot ( 1957) refers to parents still deeply preoccupied with deaths in the family that had occurred years earlier. Almost everyone practising family psychiatry who is alive to the issue will have met with similar cases.
Yet there are also cases in which the wish-fulfilment theory is certainly applicable. An example from my own experience is the intense apprehension felt by Mrs Q that Stephen might die, which was found to be a direct reaction to her own impulse to throw her baby out of the window, an impulse she had been both wholly conscious of and horrified at. What Mrs Qhad not been aware of was that her hostility to Stephen arose, in all probability, from her having redirected (displaced) towards her infant angry feelings engendered in her initially by the way her own mother behaved.
Parents of School-refusing Children: Results of Psychiatric Examination
In view of all that has been said it will come as no surprise to find, when a sample of parents of school-refusing children is examined psychiatrically, that the incidence of psychiatric disturbance is high and that, with the exception of the least severe cases, marital disharmony is universal.
Of fifty mothers studied by Hersov ( 1960b) eight had had previous psychiatric treatment (five for depressive and three for hysterical conditions) and a further seventeen were found to
215
suffer from anxiety and depression of marked degree. Of the series of thirty mothers studied by Davidson ( 1961) twelve showed symptoms of depression, including two who had been hospitalized. In a series of eighteen cases of children showing marked anxiety over separation, Britton ( 1969) reports that ten mothers had been under psychiatric treatment and another six exhibited psychiatric symptoms.
The incidence of disturbance among fathers is less pronounced though by no means negligible. Of the fifty fathers
-282-
studied by Hersov ( 1960b) eight showed psychiatric symptoms: two had had severe depressions with suicidal attempts, two others had suffered less severe depression, and another four suffered from anxiety symptoms. Davidson ( 1961) reports that eleven out of thirty fathers suffered from neurotic symptoms.
In his valuable review of the literature Malmquist ( 1965) gives much evidence of a similar kind. He is insistent that the problem is one that involves the whole family and he protests at the tendency to give too little attention to the role of father.
This completes our review of what is known of the families of children who refuse to leave home to go to school. When cases are considered in the light of the four patterns of family interaction described, it is seen, first, that, once the facts are known and the family pattern is identified, a child's behaviour is usually readily intelligible in terms of the situation he finds himself in; and, second, that many of the judgements hitherto made about such children by clinicians -- that they have been spoiled by over-indulgence, that they are afraid to grow up, that they are importunately greedy, that they wish to remain a baby tied to mother for ever, that they are fixated and regressed -- are as mistaken as they are unjust.
Two classical cases of childhood phobia: a reappraisal
In the light of our review of the family patterns that lie behind almost every case of school phobia, it is of interest to look afresh at two cases of childhood phobia that, reported during the first quarter of this century, have shaped all later theorizing. In the tradition of psychoanalysis, the classical case is that of the five-year-old Little Hans, described by Freud ( 1909). In the tradition of learning theory, a classical case is that of Peter, aged two years and ten months, described by Mary Cover Jones ( 1924b), a student of Watson.
In view of the key role that anxious attachment is held in this work to have played in all the cases of childhood phobia so far considered, is there evidence, we may ask, of its having played a part also in either of these famous cases? In what follows it is argued that in both cases there is clear presumptive evidence that it did and that, because theoretical expectations led each researcher to attend to other aspects of the case, those aspects on which weight is placed here were either overlooked or relegated to a subordinate position.
In both children the presenting symptom was an animal -283-
216
phobia. The pattern of family interaction present in the first case is likely to have been pattern B, and that in the second case pattern C.
The Case of Little Hans
A key paper in the development of psychoanalytic theorizing is Freud's study of a horse phobia in a five-year-old boy. The theory that Freud advances in that paper ( 1909) is that Little Hans's fear of being bitten by a horse had resulted from the repression and subsequent projection of his aggressive impulses, comprising hostility directed towards his father and sadism directed towards his mother. Later, he concluded: 'The motive force of the repression was fear of castration' ( 1926a, SE 20: 108). Although the origin of the hostility, oedipal or preoedipal, may have been debated by other analysts, the outline of the theory has persisted and remains the basis for all later psychoanalytic theorizing about phobias. What evidence, we may now ask, is there that anxiety about the availability of attachment figures was playing a larger part in Hans's condition than Freud realized? When the case report is read afresh in the light of our discussion of school refusal, it seems probable that anxious attachment was indeed contributing a great deal to Little Hans's problem. Most of his anxiety, it is suggested, arose from threats by his mother to desert the family. This view is advanced on two grounds:
--the sequence in which symptoms developed and statements made by Little Hans himself ( SE 10: 22-4)
--evidence in father's account that mother was in the habit of using threats of an alarming kind to discipline the boy and that those included threats to abandon him ( SE 10: 44-5).
Although the title of the paper is the 'Analysis of a Phobia in a Five-year-old Boy', Freud himself saw the child only once and the 'analysis' was conducted by Hans father. The published paper comprises father's stenographic protocol, with a running commentary and a long concluding discussion by Freud.
The parents themselves had for some years been supporters of Freud, were in fact among his first (Jones 1955), and Freud had treated mother for a neurosis before her marriage. There was a younger sister, Hanna, born three and a half years after Hans and of whom he was jealous.
Hans was four and three-quarters when father became worried
-284-
about him and consulted Freud. The problem as presented was Hans's fear that a horse would bite him in the street. Father recounted how a few days earlier Hans had been out to Scho? nbrunn with his mother, which he usually enjoyed. On this occasion, however, he had not wished to go, had cried, and had been frightened in the street when going there. On the return journey 'he said to his mother, after much internal struggling: "I was afraid a horse would bite me. "' That evening before bedtime he had remarked apprehensively: 'The horse'll come into the room. '
The symptoms, as might be expected, had not come out of the blue. According to father's record, Hans had been upset throughout the preceding week. It had begun when Hans had woken up one morning in tears. Asked why he was crying he had said to his mother: 'When I was asleep I thought you were gone and I had no Mummy to coax with. ' ( Coax was Hans expression for cuddle. ) Some days later his nursemaid had taken him to the local park, as usual. But in the street he had begun to cry and asked to be taken home, saying that he wanted to 'coax' with his mother. When later that day he had been asked why he had refused to go any further, he would not say. During that evening he had again become very frightened and cried,
217
and he had demanded to stay with his mother. The next day, his mother, eager to find out what was wrong, had taken him on the visit to Scho? nbrunn, when the horse phobia was first noticed.
Looking back further into the history, we find that the week preceding the onset of the phobia had not been the first time that Hans had expressed fear that his mother might disappear. Six months earlier, during the summer holiday, he had made remarks such as 'Suppose I was to have no Mummy' or 'Suppose you were to go away'. Looking further back still, Hans's father recalled that, when Hanna was born, Hans, aged three and a half, had been kept away from his mother. In father's opinion, Hans's 'present anxiety, which prevents him leaving the neighbourhood of the house, is in reality the longing for [his mother] which he felt then'. Freud endorses that opinion and describes Hans's 'enormously intensified affection' for his mother as 'the fundamental phenomenon in his condition' ( SE 10: 24-5; also 96 and 114 ).
Thus, both the sequence of events leading up to the phobia and Hans's own statements make it clear that, distinct from and preceding any fear of horses, Hans was afraid that his mother
-285-
might go away and leave him. Since, in the light of present knowledge, the expression of such fear should alert to the possibility that mother might have uttered threats, explicit or implicit, to leave the family, it is of interest to ask whether there is any evidence of her having done so.
Early in the record it becomes apparent that mother is inclined to use rather alarming threats. For example, when Hans is only three she is described as having threatened him that, if he touched his penis, she would send for the doctor to cut it off ( SE 10: 7-8). And we know too that, over a year later, at the time the phobia was first reported, mother was still trying to break him of the habit (p. 24 ). She is said to have 'warned him' not to touch his penis, though we are not told the nature of the warnings she was then uttering.
Three months later, however, and buried deep in the 'analytic' record, Hans lifts the curtain. He had come into father's bed one morning and in the course of talk had told his father: 'When you're away I'm afraid you're not coming home. ' Father expostulates: 'And have I ever threatened you that I shan't come home? ' 'Not you,' retorts Hans, 'but Mummy. Mummy's told me she won't come back. ' Father concedes the point. 'She said that', he replies, 'because you were naughty. ' 'Yes,' assents Hans ( SE 10: 44-5).
In the passage following father reflects, reasonably enough: 'His motive for at the most just venturing outside the house but not going away from it, and for turning round at the first attack of anxiety when he is half-way, is his fear of not finding his parents at home because they have gone away. ' Soon after, however, father reverts to an explanation along oedipal lines.
Even Hans's expressed fear that a horse might bite him is consonant with the view that fear of mother's departure was the principal source of his anxiety. This is shown by an incident that had occurred during the summer holiday of the previous year and to which Hans referred, by way of refutation, when father was trying to reassure him that horses do not bite. When Lizzi, a little girl who was staying in a neighbouring house, had gone away, the luggage had been taken to the station in a cart pulled by a white horse. Lizzi's father was there and had warned her: 'Don't put your finger to the white horse or it'll bite you' ( SE 10: 29). Thus, we find
218
Hans's fear of being bitten by a horse is closely linked in his mind to someone's departure. There is other evidence also that horses are identified with departures (e. g. p. 45 ).
-286-
On all these issues, it is evident that Freud was thinking along lines very different from those proposed here. Hans's insistent desire to remain with his mother is seen, not in terms of anxious attachment, but as the expression of his love for his mother, held to have been genitally sexual in character, having reached an extreme 'pitch of intensity' ( SE 10: 110-11). The dream that his mother had gone away and left him is held to have been, not an expression of Hans's fear that his mother would carry out a threat to desert the family, but an expression of his fear of the punishment due to him for his incestuous wishes ( SE 10: 118). The episode when Hans heard a neighbour warn that the white horse might bite is linked to a postulated wish that his father should go away, not to a fear lest his mother desert. Mother's displays of affection to Hans and her allowing him to come into bed with her are seen, not simply as a natural and comforting expression of motherly feeling, but as actions that might have encouraged, in a rather unfortunate way, Hans oedipal wishes ( SE 10: 28).
A tailpiece that tends to support the present hypothesis is that, subsequent to these events, Hans parents separated and later divorced ( SE 10: 148). (The fact that Hans was separated from his younger sister suggests that mother may have kept the little girl with her and left Hans with his father. )
There the matter must be left since there is no way of knowing which of the alternative constructions is nearer the truth. In the light of the evidence, both from the case itself and from other cases of childhood phobia reviewed earlier, the hypothesis advanced here seems no less plausible than the one adopted by Freud: it is not implausible to believe that the presenting symptom in the case of Little Hans can best be understood in terms of family interaction of pattern B.
The Case of Peter
In the literature on behaviour therapy the case of another young child, Peter, aged two years and ten months, who also suffered from intense fear of animals, has achieved some fame because it is the first recorded example of fear being deconditioned. Although the therapist, a student of Watson, assumes that the child had come to be afraid of animals through having been conditioned to fear them at some time unknown, explicit evidence regarding the way his mother treated him suggests that threats from his mother had probably played a principal part.
'When we began to study him', writes Mary Cover Jones
-287-
( 1924b), 'he was afraid of a white rat, and this fear extended to a rabbit, a fur coat, a feather, cotton wool etc. , but not to wooden blocks and similar toys. ' At the sight of a white rat in his crib ' Peter screamed and fell flat on his back in a paroxysm of fear', and he proved to be even more afraid of a rabbit. Since other children of the same age were not particularly afraid of these creatures the researchers decided to see whether they could help Peter to become less afraid of them.
219
A principal procedure used to 'decondition' Peter was for him to play each day with three other children chosen because of 'their entirely fearless attitude toward the rabbit'; during a part of the play period the rabbit was brought in. After about nine sessions, a second procedure was added: each time before the rabbit appeared Peter and the other children were given candy. Altogether some forty-five sessions were given, strung out over a period of nearly six months, during which there was a two-month interruption while Peter was in hospital for scarlet fever. From time to time Peter's progress was tested by presenting him with the rabbit when he was alone. At the end of the process Peter was no longer afraid of the rabbit or of the feather and he was much less afraid of the rat and the fur coat.
From the viewpoint of this work two aspects of the case command attention.
First, Peter is described as having come from a disturbed family that was living in impoverished conditions. Throughout the experiment, it seems, he was in a residential nursery 1 or else a hospital. His mother is described as 'a highly emotional individual who cannot get through an interview without a display of tears'. Peter's older sister had died and the parents were said thenceforward to have lavished 'unwise affection' on him. Discipline was 'erratic' and in her attempts to control him mother is said to have resorted to threats: the example given is 'Come in Peter, someone wants to steal you'. The pattern of family interaction suggested by this limited information is pattern C.
The second point of interest is the effect on the deconditioning process of the presence or absence of a particular student assistant of whom Peter was fond and who he insisted was his father.
____________________
1 Although it is not stated explicitly that Peter was resident in the nursery in which
deconditioning took place, a phrase (near the end of the paper) that 'he has gone home to a difficult environment' suggests that this was so.
-288-
On each of two occasions when this assistant was present Peter became decidedly less afraid, although the assistant made no overt suggestions. On this phenomenon Jones comments, 'it may be that having him there contributed to Peter's general feeling of well-being and thus indirectly affected his reactions'.
Animal phobias in childhood
There is no disposition to argue here that every case of animal phobia in childhood and later life is but the tip of an iceberg the great bulk of which comprises intense fear of losing an attachment figure. In some individuals, no doubt, an animal phobia has developed because as children they had some frightening experience in which they were attacked by an animal of that species. In other cases, seeing or hearing about such events, perhaps in dramatic circumstances and at an age when misunderstanding and fallacious over-generalization are common, may be responsible. In yet others, prolonged exposure to a parent or other adult who habitually responds with fear to a particular species of animal may play a part. Whatever the causes, Marks ( 1969) presents evidence suggesting that there are individuals who are acutely afraid of a particular species of animal but who do not suffer from any other form of emotional disturbance.
220
Yet, although cases of true and limited animal phobia may well exist, there can be little doubt that, in very many children and probably adults also who are more than usually afraid of animals, the principal source of anxiety lies in the home and not outside it. It is already suggested that the cases of Little Hans and Peter can usefully be considered in that light. Further and substantial evidence stems from the finding that, as already described, many school-refusing children number among their heterogeneous symptoms a fear of animals. Furthermore, just as any expressed fear of school sinks into oblivion once the disturbed family situation is recognized and dealt with, so does any expressed fear of animals. Because that is so and because difficulties at home are so frequently kept a secret, it is wise when confronted by a patient complaining of animal phobia always to examine carefully the pattern of interaction within the family from which the patient comes.
The wisdom of this course is well illustrated by a case of animal phobia in an adult reported by Moss ( 1960). The patient was a woman of forty-five who had suffered since childhood from an intense fear of dogs. After seeing a film ( The Three Facesof Eve
-289-
of Eve) in which a woman is treated for phobia by means of hypnosis, she sought hypnotic treatment for herself.
During the course of treatment the patient recalled a tragic event that had occurred when she was aged four. It appeared that she had been playing in the backyard of her home with her younger sister when the family dog, Rover, knocked the little sister down. A splinter entered the child's cheek, the wound went septic and a few days later the child died. The patient recalled how her mother had accused her of having knocked her sister down and had openly blamed her for the death, and also how thenceforward she had deeply disliked Rover and had become afraid of dogs of every kind. A few years later, after another sister had been born, she recalled that she had been much afraid lest a puppy attack that sister also.
After the patient had recalled how her mother had blamed her for her sister's death, much in her life seemed to her to fit into place. For example, the episode seemed to explain, at least in part, why she had always felt misunderstood by her mother, why she had suffered from a chronic sense of guilt and a compulsive desire to please, and why her relationship to her mother had been so deeply ambivalent.
When an event that a patient recalls has occurred many years previously it is extremely difficult to be sure how valid the recalled details may be. In this case it was possible to obtain limited corroboration of the patient's story. An elder brother confirmed the presence of Rover and also that at the time of the fatal accident his two little sisters had been alone, because he and his brother, who seem to have been left in charge, had gone off to watch a fire. The patient's youngest sister recalled how in later years the patient had anxiously protected her from the approach of any and every dog. There was, however, no corroboration that mother had blamed the patient for the accident, and mother herself, who was still alive, denied having done so.
Experience in family psychiatry shows, nevertheless, that, when a young child dies, it is by no means uncommon for a parent, distraught by what has happened and perhaps feeling guilty over failure to have taken some precaution, impetuously to attribute blame to an older child. In some families the older child then becomes a scapegoat; in others the parent, after
221
recovering from the shock of acute grief, may forget, and then deny, having ever made the accusation. But in either case the accusation cuts deep, even when memory of it is repressed.
-290-
That is what seems likely to have occurred in the case described. If that were so, there would be little wonder that the accused child had come to hate and fear the animal that she believed responsible for her disgrace. Nor would there be wonder that she should have felt thenceforward that her mother, and therefore all others to whom she might look for comfort and support, would disown her and treat her with nothing but contempt.
Enough has been said perhaps to show that the theory of anxious attachment outlined in earlier chapters can illuminate many a case in which a child is intensely and persistently afraid of some situation in circumstances that are perplexing to all around him and perhaps also to the child himself. In the next chapter the problem of agoraphobia in adults is considered in the light of the same theory.
-291-
Chapter 19
Anxious Attachment and 'Agoraphobia'
It follows from the nature of the facts . . . that we are obliged to pay as much attention in our case histories to the purely human and social circumstances of our patients as to the somatic data and the symptoms of the disorder. Above all, our interests will be directed towards their family circumstances . . .
SIGMUND FREUD ( 1905a)
Symptomatology and theories of 'agoraphobia'
When a psychiatrist used to dealing with children and families examines the problem of 'agoraphobia' 1 he is at once struck by its resemblance to school phobia. In both types of case the patient is alleged to be afraid of going into a place filled with other people; in both the patient is apt to be afraid of various other situations as well; in both the patient is prone to anxiety attacks, depression, and psychosomatic symptoms; in both the condition is precipitated often by an illness or death; in both the patient is found to be 'overdependent', to be the child of parents one or both of whom suffer from long-standing neurosis, and frequently also to be under the domination of an 'overprotective' mother. Finally, a significant number of agoraphobic patients were school refusers as children.
Although minor degrees of agoraphobia are probably common and, when of recent origin, probably have a high remission rate ( Marks 1971), patients who come to the attention of psychiatrists are usually those who are suffering either from a chronic condition of some severity or else from an acute attack. Often a patient is intensely anxious, apt to panic when unable
____________________
1 The condition under discussion appears in the literature under many names, including
222
anxiety hysteria, anxiety neurosis, anxiety state, and phobic anxiety-depersonalization syndrome ( Roth 1959). The name most widely adopted at present is agoraphobia ( Marks 1969). Since criteria used to select cases differ from study to study, the extent to which findings are comparable remains in doubt.
-292-
to get home quickly, and to be afraid of an extraordinarily broad range of situations (typically, crowded places, the street, travelling) or of collapsing or even dying when out on his own. From among this heterogeneous and variable collection of situations feared it is possible, none the less, to identify two that are feared in virtually every case and are also the most feared. These situations are, first, leaving familiar surroundings and, second, being alone, especially when out of the house. Since the argument advanced here turns on the fact that fear of these situations is at the heart of the syndrome, let us consider the evidence.
During the past decade there has been very active interest in the syndrome by psychiatrists in the United Kingdom. Roth and his colleagues in Newcastle upon Tyne describe two series of cases, each numbering over one hundred ( Roth 1959; 1960; Harper & Roth, 1962; Roth, Garside & Gurney 1965; Schapira, Kerr & Roth 1970). Special aspects of the condition to which they give attention are: the high incidence of traumatic precipitating events, notably actual or threatened physical illness, bereavement and illness in the family; the high incidence of depersonalization; and the close relation of the condition to states of anxiety and depression. Another programme of research into the condition, with special reference to the efficacy of different methods of treatment, is one conducted at the Maudsley Hospital, London, by Marks and Gelder (for references to their numerous papers see Marks 1969 and 1971). A third study of value is by Snaith ( 1968) who reports on forty-eight cases of phobia in adult patients, twenty-seven of whom were typically agoraphobic. Roberts ( 1964) describes results of a follow-up of thirty-eight patients, all married women.
Although none of these workers approaches the problem from a standpoint in any way similar to that adopted here, each endorses the view that a principal feature of the condition is fear of leaving home. Roth ( 1959) speaks of 'a fearful aversion to leaving familiar surroundings'; Marks ( 1969) holds that 'fear of going out is probably the most frequent symptom from which others develop'; Snaith ( 1968) finds that, in twentyseven of his forty-eight cases, the principal source of fear is leaving home and its attendant circumstances. Furthermore he reports, first, that the more anxious an agoraphobic patient becomes the more intense grows his fear of leaving home and, second, that when a patient becomes more anxious his fear of leaving home is magnified in intensity by a factor many times
-293-
greater than is his fear of anything else. These findings lead Snaith to suggest that the condition is not a true phobia and that a more appropriate label for it would be 'non-specific insecurity fear'. In keeping with Snaith's perspective is the criterion that Roberts ( 1964) laid down for inclusion in his series, namely a patient's inability to leave his house without a companion.
Not only do these workers find that fear of leaving home when unaccompanied is the principal feature of agoraphobia, but each of them reports also that most patients have been anxious individuals all their lives: some for decades have been uneasy about going out alone (
223
Marks 1969). Between 50 and 70 per cent of patients are reported to have suffered from fears and phobias during their childhood ( Roth 1960; Roberts 1964; Snaith 1968). In a recent survey of 600 cases using a questionnaire, between one-fifth and one-sixth described themselves as having been in some degree 'school phobic' ( Berg, Lipsedge & Marks, in preparation).
Again, although psychoanalysts working in the classical tradition have an approach to the problem entirely different from that of any of the workers so far cited, and different also from that adopted in this work, they report almost exactly the same findings. For example, in an early paper describing the case of a small boy, Abraham ( 1913) notes that the boy 'does not speak of fear, but of his desire to be with his mother'. This leads Abraham to conclude that the basic problem in patients suffering from agoraphobia is that their 'unconscious . . . does not permit them to be away from those on whom their libido is fixated'.
Both Deutsch ( 1929) and, in recent years, Weiss ( 1964) endorse Abraham's view. Weiss notes especially that a patient's anxiety is apt to increase the further from home he goes, which leads him to define agoraphobia as 'an anxiety reaction to abandoning a fixed point of support'.
Thus, despite great variation in the approach and outlook of these many workers, the findings they report are impressively consistent. Only when attempts are made to accommodate the findings within a theoretical framework do differences and difficulties arise.
Three Types of Theorizing
Here, as so often elsewhere, the two rival types of theory that dominate the field are psychoanalytic theory and learning
-294-
theory. In the case of agoraphobia, however, a third type of theory has also been advanced, namely Roth's psychosomatic theory which invokes both psychological and neurophysiological processes ( Roth 1960). Strikingly enough, despite all the telltale hints that a major part is played by relationships within the patient's family of origin, a fourth type of theory, namely one that invokes pathogenic patterns of family interaction as major aetiological agents, is conspicuous by its absence.
1. Psychoanalytic theories of agoraphobia come in two main variants according to whether they focus on fear of being in the street or fear of leaving home.
Freud tends to concentrate on fear of being in the street, which he sees as a displacement outward of the patient's fear of his own libido. Even though in 1926 Freud began a major revision of his views and reached the conclusion that 'the key to an understanding of anxiety' is 'missing someone who is loved and longed for' (see Chapter 2 of this volume), he never applied his new theory to agoraphobia. 1 As a result, his original hypothesis continues to be invoked by a number of psychoanalysts who still see sexual temptation, of one kind or another, as the principal situation that an agoraphobic patient fears (e. g. Katan 1951; Friedman 1959; Weiss 1964).
224
Other psychoanalysts in their theorizing take as their focus a patient's fear of leaving home and, in doing so, advance theories very similar to those their colleagues advance to account for the similar fear found in children diagnosed as suffering from school phobia. Thus Deutsch ( 1929) notes that the reason an agoraphobic patient feels compelled to remain near his mother (or other loved person) is that he entertains unconscious hostile wishes against her and so has to remain with her to ensure that his wishes are not enacted. For Weiss ( 1964) the patient's urge to remain at home is to be understood as due to a 'regression to unresolved dependency needs'. This is also the view of Fairbairn ( 1952), although in his case histories he attributes a causal role to the very insecure childhoods his patients had experienced.
In none of the psychoanalytic formulations, apart perhaps
____________________
1 In one of his last works, New Introductory Lectures ( 1933), Freud writes: 'the agoraphobic
patient is . . . afraid of feelings of temptation that are aroused in him by meeting people in the street. In his phobia he brings about a displacement and henceforward is afraid of an external situation' ( SE 22: 84).
-295-
from Fairbairn's, is there any suggestion that a patient's refusal to leave home is a response to the behaviour of one of his parents, not only behaviour that may have occurred at some time during the past but behaviour that may be occurring still in the present.
2. During the past decade or so a new approach to a theoretical understanding of phobic conditions of all sorts has been made, this time by learning theorists; and formulations that attempt to account for each of the various situations feared have been advanced. Whereas this approach may well help us to understand some of the discrete animal phobias, how much it can contribute to an understanding of agoraphobia remains in doubt. Describing the present position as he sees it Marks ( 1969), who has made a special study of agoraphobia and draws extensively on learning theory, writes as follows:
Certain phobias, especially agoraphobia, are commonly found together with multiple other symptoms such as diffuse anxiety, panic attacks, depression, depersonalization, obsessions and frigidity. Learning theory does not explain why these symptoms develop, why they occur together, nor why they are associated more often with agoraphobia than with any other type of phobia.
Furthermore, in Marks's view, 'the origin of the panics, depression and other symptoms is not indicated by learning theory' (p. 93 ).
How the panics and depressions do originate is, for Marks, the most puzzling aspect of the condition. For, in his opinion, not only is learning theory unable to account for them but no other theory can do so either (p. 93 ). Admitting the quandary, Marks leaves the matter open; but he tends to the view that anxiety attacks probably have an unknown physiological origin. Nowhere does he consider the possibility that they may originate in family situations that create psychological distress.
Having recognized frankly the difficulties in accounting for agoraphobic symptoms entirely in terms of learning theory, Marks believes nevertheless that the theory has much to offer. The
225
hypothesis he advances is based on the idea, suggested by learning theory, 'that panic attacks and depression [may] act as super-reinforcers which facilitate phobic conditioning' whenever a patient who happens to be experiencing such affects goes out of his home. This line of thought leads Marks to propose that, in the development of agoraphobia, the anxiety
-296-
attack comes first and the situations that the patient reports he fears come to be feared only later, either as a result of a secondary conditioning effect or as a result of rationalization. In that context both fear of going out of the house and fear of becoming separated from a companion, the two symptoms most characteristic of agoraphobic patients, are held to develop through a process of secondary conditioning.
In keeping with his hypothesis, Marks expresses much scepticism regarding the causal role of precipitating factors, holding that they probably act simply as 'non-specific stressors in a patient already liable to the disorder . . . or that the disorder was already present, but hidden until the stressor elicited or exacerbated it'. In support of his position he lays much emphasis on his claim that 'not a few phobias start suddenly without any obvious change in the life- situation of the patient' (p. 128 ).
Both the sequence of events that Marks postulates and the weakness of his position are illustrated in his description of the case of a woman who sought treatment at the age of thirtythree on account of depression with suicidal ideas. The account she then gave was that, ten years earlier when she was aged twenty-three, she had developed anxiety, sweating, and shaking of the legs while travelling to work by train. Subsequently she had discovered that she felt better if her husband was present and so had taken a job in the firm in which he worked. After a few months, however, she had become afraid of separation from him, had to know exactly where he was and had telephoned him frequently. If for any reason she could not contact him immediately she would panic, feel completely lost and want to scream.
The only information Marks gives regarding this patient's childhood is that 'as a small child [she] used to be frightened when her parents were out and once sent out her younger brother to find them. She had infrequent desires to scream which were hard to stifle. These disappeared in her late teens. '
Despite the uncertain validity of the retrospective data, Marks seems confident about the sequence of symptoms: 'First came the travel phobia and depersonalization, then came the discovery of relief in the presence of her husband and after this he became indispensable. Finally the patient presented for treatment of separation anxiety. ' In accounting for the symptoms Marks proposes two distinct pathologies. On the one hand is the agoraphobia and on the other is the anxiety about
-297-
separation, to which he believes the patient had been sensitized as a child. Originating independently, the two pathologies are held subsequently to have interacted.
There are several flaws in Marks's position. First, in the light of the childhood history of this patient, it is difficult to accept his confident assertion that agoraphobia came first and separation anxiety afterwards. Second, in his ready acceptance of this and other patients'
226
accounts that the initial anxiety attack came 'out of the blue', he makes no allowance for a patient's witting or unwitting suppression of information, a process we know to be extremely common and often to hide clues vital for understanding the condition. Third, to postulate two distinct psychopathologies for a pair of symptoms that habitually go together 1 is far from parsimonious. Finally, as Marks himself admits, he can give no explanation of how or why this patient (or any other) first started to experience anxiety and panic attacks.
An alternative hypothesis to account for this patient's symptoms is that, during her childhood, she had been subjected to repeated and realistic threats of being abandoned, so that even after reaching adult life she had continued to be acutely sensitive to any such danger. The case is discussed further below (p. 307 ).
As regards the part played by a patient's parents in the genesis of agoraphobia, learning theorists share with traditional psychoanalysts the same shadowy picture. Whereas neither group attributes much importance to parental behaviour, in so far as they do so both invoke the theory of spoiling. As Andrews ( 1966) points out, Wolpe ( 1958) and Lazarus ( 1960), two leading learning theorists, both regard a patient's tendencies to withdraw to and to remain at home as responses he has learnt during interaction with over-protective parents. Marks ( 1969) in his discussion of prevention implies the same process. Some years earlier Terhune ( 1949), a psychiatrist whose outlook is in many ways similar to that of present-day learning theorists, wrote confidently: 'The phobic person is one who has been over-protected, brought up "soft". '
3. The third main type of theorizing about agoraphobia,
____________________
1 Marks rests his argument on his finding that about 5 per cent of agoraphobic patients are
not helped by company and prefer to be alone while travelling ( Marks 1969: 98). in most syndromes, however, cases occur that lack one or more typical symptoms; an example is measles without a rash. Such atypical cases require special study.
-298-
initiated by Roth ( 1959; 1960), regards the condition as being truly psychosomatic. In presenting his theory, Roth lays much emphasis on the vulnerable personalities of his patients, on the precipitating role of stressful events, and on depersonalization, which he regards as a core symptom of the syndrome. The psychological factors to which he points include both situations that, acting perhaps from early childhood, are thought to have contributed to the development of an anxious dependent personality, and stressful events, such as bereavement and illness, that appear to act as precipitants. The somatic factor he postulates is a specific cerebral mechanism that, once triggered, is difficult to inactivate. After considering certain disturbances of perception and consciousness that he finds in these patients, together with symptoms that he attributes to temporal lobe dysfunction, Roth concludes that the somatic pathology is probably to be understood as arising in the mechanisms regulating awareness, which he postulates to have become chronically deranged. Although he gives little detail of how he believes a difficult childhood and stressful situations of later life interact to produce an agoraphobic syndrome, Roth's approach is not incompatible with that presented here.
We turn now to consider a fourth type of theory, namely the one that results when the problem of agoraphobia is looked at in the theoretical perspective developed in this work.
227
Throughout our further discussion it is important to bear in mind that, as all workers now agree, the central symptom of the condition under scrutiny is fear of leaving home.
Pathogenic patterns of family interaction
Whether the theory of anxious attachment, applied already to problems of school phobia, can help to solve problems of agoraphobia also must remain in doubt. For, apart from some limited and mostly rather crude observations, there are few data yet available on patterns of interaction within an agoraphobic patient's family of origin. Almost all the data yet published come either from the patient himself or from a single interview with a relative, with the many distortions and omissions that we know such clinical procedures entail. What is missing, but so necessary to have, is first-hand observations of how a patient and his parents are currently behaving towards one another. In the absence of such data all that is possible here is to draw attention to certain reasonably well-attested findings
-299-
that are at least consonant with the view that many, if not all, cases of agoraphobia can be understood as products of pathogenic patterns of family interaction. There is much evidence of a rather general kind which, although it gives little information about specific patterns of interaction, points to a high incidence of disturbance in the families from which agoraphobic patients come. Before considering specific patterns, therefore, we consider this general evidence. Most reports agree that a majority of agoraphobic patients come from homes that are intact, in the sense that there are two parents living continuously together. Yet there is also substantial evidence that, within these homes, relationships are often far from harmonious; and it is repeatedly indicated that the parents of patients are neurotic or disturbed in some other way. Taking as a criterion clear-cut neurotic breakdown in first-degree relatives, Roth ( 1959) finds an incidence of 21 per cent. Nor should it be overlooked that a minority of patients, in one study as high as 25 per cent, come from homes broken by death, divorce, or other cause. In a recent study of eighty-seven consecutive London patients by M. S. Lipsedge (unpublished), a high incidence of disturbance in the families of origin is reported. The patients ranged in age from twenty-two to sixty-four years; fourteen were men and seventy- three women. Almost all the information was obtained from the patients themselves during an initial interview, though occasionally it was supplemented by information from a general practitioner. Inadequate though this method is for obtaining the information required, it is unlikely to exaggerate the degree of disturbance in the families from which the patients came. On the basis of this information, the patients' families can be divided very roughly into three categories:
I. I intact and reasonably stable families
II. II intact families in which there was much quarrelling, violence, and alcoholism, and/or
almost complete absence of affection
III. III families broken by death or divorce, or in which a parent was chronically ill, and/or
the patient had had prolonged separations from or changes of parent figure.
The number and proportion of patients from families in each category are shown in the following table.
-300-
228
Category of No.
of % of patients
42 30 28 100
family
I
II
III Total
patients
37 26 24 87
Category I: There were thirty-seven patients who described their home life as having been happy or who gave no particularly adverse information about it. Nevertheless, two of these patients described a parent as having been agoraphobic (one father and one mother), and two others described themselves as having been 'over-protected'. Ten of the patients referred to themselves as having been notably fearful as children; of these, two had been school refusers and one agoraphobic. Thus neurotic trouble of one kind or another is reported in members of about one-third of these not overtly disturbed families.
Category II: There were twenty-six patients whose homes appear to have been intact but who described themselves as having had an extremely unhappy family life when children. Eighteen patients described their parents as having engaged in perpetual quarrels, including violence, and often made worse by alcohol. Another eight complained of having received no affection and/ or of having been rejected. In three of these twenty-six cases the patient's mother had been agoraphobic. Two of the patients had themselves been school refusers as children.
Category III: Of the remaining twenty-four patients, there were twenty-one whose family life had been disrupted by death, divorce, or desertion and/or who had experienced many changes of mother figure. Of these, ten had lost one or both parents by death before their tenth birthday (six a father, three a mother, and one both). In five cases mother had deserted and in at least one other father.
