A child is instructed to put his head on the seat of one chair and his feet on that of the other and to stiffen his back to make a bridge, in which
position
he must remain.
Bowlby - Separation
In the situation described:
--sight of opera glasses and sound of whistle together were not alarming, nor was there reason to think that either singly would be
--sight of the caterpillar aroused interest which alternated with slight fear
--sight of caterpillar and sound of whistle together were alarming.
From observations such as these Valentine suggests that there is a much readier bias to develop fear of objects like caterpillars than fear of objects like opera glasses. 1
The extreme ease with which monkeys and apes develop a fear of snakes is remarked on above (Chapter 8, p. 130). The same is true of humans. As already reported in Chapter 7 (pp. 109-13), in the experiments conducted by Jersild & Holmes ( 1935a) between one-third and a half of the children aged between two and six years showed marked fear of the snake. A comparable finding is reported by Morris & Morris ( 1965).
____________________
1 A defect of this experiment was that during the second test when the caterpillar was
presented and the whistle blown Y was seated on her father's knee, not her mother's. It is therefore possible that a change in the person caring for her was responsible for the results.
-163-
132
In a children's television programme in England children were invited to compete for a prize by proposing future programmes. To qualify, however, they had to name the animal they liked most and the animal they disliked most. Altogether nearly twelve thousand children aged from four years upwards replied. Of the animals most disliked the snake was an easy first: it was named by 27 per cent of the children. Spiders came next, given by under 10 per cent; then lions and tigers, together, were named by about 7 per cent. Up to the age of nine years, at least one child in three expressed dislike of snakes. At all ages slightly more girls than boys expressed the fear.
It seems likely that to the development of a fear of animals in general and of snakes in particular several factors make an interlocking contribution. First are several of the common natural clues, including often strangeness. Second, there may also be certain specific natural clues, for example crawling or wriggling. Third, there is the behaviour of others. Because of their appearance and behaviour, including their vocalizations, animals arouse simultaneously both lively interest and incipient fear. In such conditions the behaviour of a companion is likely to have a maximum effect, tipping the balance either towards decreased fear and approach or towards increased fear and withdrawal.
Fear of Darkness
Every study shows that fear of darkness is as common at every age as fear of animals and that during ontogeny it runs a roughly parallel course. In all likelihood the development of fear of darkness is to be explained in a comparable way to that of fear of animals; though the natural clues concerned are usually not the same ones.
In conditions of darkness the two natural clues that are apt to be present together are strangeness and being alone. During darkness visual stimuli that, if viewed in daylight, would be recognized as familiar are often ambiguous and difficult to interpret. Endless examples come to mind: the pattern of movement of light shining through bedroom curtains; the shapes of the trees in a wood at night; the shadowy recesses of a dimly lit cellar. In each case the visual stimuli available are barely adequate for accurate perception and it is therefore as easy to perceive something unusual as something familiar. In addition, without visual cues, sounds are far more difficult to interpret accurately or with confidence. Thus in conditions of
-164-
darkness much seems uncertain or strange and in consequence alarming.
Yet, mere strangeness would probably of itself arouse comparatively little fear were it not so regularly accompanied by being alone. Sometimes a person is in reality alone; sometimes, because his companion is unseen, he may merely feel himself to be. In either case the situation is compound: it combines sights and sounds that are not easily interpreted and the situation of being alone.
Freud, it is interesting to note, was greatly struck by the way in which darkness leads a child to feel alone, and it is an observation of a small boy's behaviour in the dark, together with inferences from it, that lies at the heart of his theory of anxiety. This is therefore a good point at which to compare Freud's theory with the one presented here.
133
In Three Essays ( 1905b, SE 7: 224n) and again in the Introductory Lectures ( 1917b, SE 16: 407) Freud tells the story of a three-year-old boy. He recounts how he once heard this child
calling out of a dark room: 'Auntie, speak to me! I'm frightened because it's so dark. ' His aunt answered him: 'What good would that do? You can't see me. ''That doesn't matter,' replied the child, 'if anyone speaks, it gets light. ' Thus [comments Freud] what he was afraid of was not the dark, but the absence of someone he loved . . .
Reflection on that episode, Freudtells us, led him to the view that the prototypic situation that gives rise to anxiety in children is simply separation from mother. Neurotic anxiety, he then argues, can best be understood as a persistence beyond childhood of the tendency to be anxious when alone, though fear of being alone often masquerades as fear of something else, for example of the dark. In all these regards the theory advanced here is very close to Freud's. Where the two differ is that Freud did not recognize that strangeness is intrinsically frightening or that both strangeness and being alone can usefully be regarded as two members of a class of natural clues to increased risk of danger. As a consequence he held that to be afraid when alone (and also when confronted by any of the other natural clues) is irrational and neurotic; whereas in the theory put forward here to be afraid in such conditions is held to be in general adaptive.
-165-
Fear of Being Alone
Repeatedly in these chapters it is stressed that being alone is one of several natural clues to increased risk of danger and that it occurs very commonly as a component in a compound situation. Not only does it occur in combination with other natural clues, moreover; it can occur equally well with cultural clues and also in situations realistically assessed as being dangerous. Thus, throughout life, being alone is a condition that either stimulates fear or greatly intensifies fear aroused in other ways. Concomitantly, being with a companion greatly reduces fear. In no conditions is the reassuring effect of the presence of companions more evident than during and after a disaster.
Behaviour in disaster
The role of a companion in reducing fear in children is very obvious and it is also readily acknowledged by children. Adults, by contrast, are less likely to acknowledge it. During and after a disaster, however, people are less reticent ( Baker & Chapman 1962).
When the impact of disaster comes members of a family commonly cling together:
When sirens scream of approaching disaster, minds turn to loved ones. If they are near enough mothers run to protect their children, and men seek their families. They huddle together and support one another through the stress, and when it is passed they resume and nurse those they love ( Hill & Hansen 1962).
Wolfenstein ( 1957) describes how a woman who had been with her fifteen-year-old daughter when a tornado struck recounted her experience:
134
And she said, 'Mother, its coming -- a cyclone. ' And I said, 'Mary, I'm afraid it is. But,' I said, 'we're together. ' And she said, 'Mother, I love you and we're together. ' I shall never forget those words. And we -- our arms were around each other, and I said, 'Whatever happens, Mary, let's cling together. '
Whenever members of a family happen to be apart at the time of impact they are unlikely to rest until they have found each other; then, again, physical embraces are the rule. 'Just
-166-
being together is deeply important following impact even in loosely-knit families' ( Hill & Hansen 1962).
Survivors agree that to be alone during a disaster is extremely frightening, whereas the advent of a companion, however inadequate, is likely to transform the scene. Wolfenstein refers to another episode, one in which, following an explosion, two injured men were trying to crawl out of a burning factory. Describing their experiences, one of the men, who had suffered a broken leg, explained:
Then Johnny and Clyde came along. I said, 'Johnny, help us -- we can't walk. ' His arms were broken and he said, 'I can't help you, but I'll stay with you. If you can crawl, I'll guide you. ' Talk about cheer! That helped me more than anything -- just when he said, 'I'll stay with you. '
Not only is there a strong tendency for members of a family or other social group to remain together during the height of a disaster but the tendency is likely to persist for days or weeks after it is over. This heightened tendency to attachment behaviour is commented on in a number of reports.
For example, Bloch, Silber & Perry ( 1956) studied the effects on children of a tornado that struck a town in Mississippi, affecting in particular a cinema in which children were attending a Saturday afternoon programme. Altogether, information was obtained during the succeeding weeks from interviewing the parents of 185 children between the ages of two and twelve years.
About one-third of the children were reported to be showing signs of increased anxiety, which typically took the form of clinging to or remaining close to parents and wishing to sleep with them. They were made anxious by noise and also tended to avoid situations associated with the tornado. Children aged six to twelve were more disturbed than younger ones. A possible reason for that was that more of them had probably been in the impact zone. Another possible reason, though it is not commented on by the researchers, was that the older the child the more likely was he to have been away from his parents. Boys were as much affected as girls.
Experiences that were significantly associated with increased anxiety were the child's presence in the impact zone, personal injury, and the death or injury of a family member. Not unexpectedly the reaction of children reflected the reaction of parents. In nine cases parents described themselves as having
-167-
135
'gone to pieces' and, instead of having supported their child, as having sought help from him. Eight of these children were disturbed and, about the ninth, the mother could not be induced to talk. Further discussion of the contribution to a child's anxieties of parents who invert the relationship by requiring their child to care for them will be found in Chapters 18 and 19. Many cases diagnosed as school phobia and agoraphobia can be understood as being caused by such inversions.
Reports on the effects of the Mississippi tornado of 1953 ( Blochet al. 1956) and of the Los Angeles earthquake of 1971 ( Time, 8 March 1971) both make it clear that, after a disaster, parents are almost as eager to retain their children close to them as the children are eager to remain close to their parents. Since these responses are adaptive, it is unfortunate that the concept of regression is so frequently invoked to explain them. Investigation shows that, in run-of-the-mill cases as well as in those that follow a disaster, behind behaviour dubbed regressive by clinicians there exist situations that, once known about, explain at once why a child or an adult should cling relentlessly to another member of his family.
-168-
Chapter II
Rationalization, Misattribution, and Projection
All round the house is the jet-black night;
It stares through the window-pane;
It crawls in the corners, hiding from the light, And it moves with the moving flame.
Now my little heart goes a-beating like a drum,
With the breath of the Bogie in my hair;
And all round the candle the crooked shadows come, And go marching along up the stair.
The shadow of the balusters, the shadow of the lamp, The shadow of the child that goes to bed --
All the wicked shadows coming, tramp, tramp, tramp, With the black night overhead.
ROBERT LOUIS STEVENSON, A Child's Garden of Verses
Difficulties in identifying situations that arouse fear
When a person is afraid and claims that something in particular, for example thunder or a dog, has made him so, doubt is often expressed whether he has identified the right stimulus situation. This is especially likely to happen when fear is shown or reported by children and by emotionally disturbed adults. Among psychoanalysts there is a long tradition of claiming that what a person is really afraid of is something very different from what he claims to be afraid of. Indeed, psychoanalytic theorizing about anxiety and fear reflects a prolonged hunt for some primal danger situation that is thought to arouse a primal anxiety or fear. 1 Arising out of that tradition, also, is the practice of invoking the process of projection whenever a fear appears not to be appropriate to the situation presenting.
136
In the approach adopted here, no less than in the traditional ones, misattribution is held to be very common. The difference in the present approach lies in the explanations it offers of why misattribution should occur. The concept of a primal danger
____________________
1 See Chapter 5 and Appendix I.
-169-
situation is dispensed with, and projection is given a much smaller role as an explanatory principle. A solution is found in the relationship that the natural clues bear to danger and safety.
The very fact that fear is first aroused in human beings not by any rational appraisal of danger but by stimulus situations that are no more than clues to an increased risk of danger invites misunderstanding and misattribution; for, as has been made clear, a natural clue is in no sense inherently dangerous. Because that is so, however, and because in Western culture (and perhaps also in others) a human being is expected to be afraid only of real dangers, there is a strong bias both in the frightened subject and in an onlooker to attribute the fear response to something other than the natural clue. For example, since it is thought absurd for anyone to be afraid merely of thunder, the fear is 'explained' as a fear really of being struck by lightning. Similarly, since it is thought absurd to be afraid merely of a dog, the fear is 'explained' as a fear really of being bitten by a dog.
Rationalizations of these sorts are no doubt very common. They are commented upon by all who have made a study of fear, irrespective of theoretical orientation. For example, Marks ( 1969) suggests that a child's fear of monsters in the dark may be no more than a rationalization of his fear of the dark, 'a genuine rationalization of an irrational fear on the same lines that any post-hypnotic suggestion is rationalized'. The Newsons ( 1968) point out that such rationalizations are often and easily encouraged by other children, or even by adults, who tease a child about what he might meet when in the dark on his own. Jersild ( 1943) calls attention to the fact that when a child is already frightened, from whatever cause, he 'may formulate his fear in terms of an imaginary or anticipated danger', such as criminal characters or bogeys or some other sinister circumstances he has encountered or, more probably, heard or read about.
Though simple rationalizations of this kind are probably common, even commoner perhaps are mistaken or biased attributions stemming from the special properties of compound situations. In compound situations, two or more stimulus conditions, by being present together, have the effect of arousing far more intense fear than would any one were it to occur separately. In such a case there is a marked tendency to single out one component of the compound situation as the one that arouses the fear and to ignore the other(s). For example, a
-170-
person is alone in the dark and hears strange noises. Whereas in fact all three conditions -- being alone, being in the dark, and hearing the strange noises -- may well be necessary to account for the fear aroused, in all likelihood attention is focused solely on the strange noises and the other components of the situation are almost ignored. Furthermore, from there it is only a short step to rationalize the fear, aroused in fact by what is little more than a
137
combination of two or three natural clues, and to claim that what is feared is burglars or ghosts.
Which one of several components present in a compound situation is fastened on as the fear- arousing one and which are neglected needs examination. Presumably, the component selected is usually the one that most readily lends itself to being interpreted as indicative of real danger. If that is so, being alone would habitually be neglected, or at least given a subordinate position. That is, in fact, very close to what Freud believed to happen, though he expresses his views in terms of libido theory and not of attachment theory.
In 1917, at the end of a discussion of the psychopathology of phobias, Freud summed up his position:
Infantile anxiety has very little to do with realistic anxiety, but, on the other hand, is closely related to the neurotic anxiety of adults. Like the latter, it is derived from unemployed libido, and it replaces the missing love-object by an external object or by a situation ( 1917b, SE 16: 408)
Since Freud regards the unemployed libido as constituting an internal danger, his formulation is that fear of an internal danger is replaced by fear of an external one. An alternative rendering of his position would run: when a child or an adult is afraid of some external object or situation, what he is really afraid of is the absence of someone he loves.
In Chapters 18 and 19, in which misattributions are discussed further, reasons are given for believing that many intense fears attributed to all sorts of common situations and termed phobias are best understood as being aroused in compound situations, a main component of which is the expectation of being separated from a principal attachment figure. The famous and theoretically influential case of 'Little Hans', who was afraid of being bitten by a horse, appears to be a good example ( Freud 1909, SE 10). Evidence is presented (Chapter 18) for believing that fear of separation played a much larger part in this case than Freud at the time realized.
-171-
Misattribution and the role of projection
In some psychoanalytic traditions the concept of projection has been used very extensively in an attempt to explain any fear that is not readily intelligible as a response to a real danger. Since the term is itself used in several ways the resulting theory is often confused.
One usage of projection is to denote our propensity to perceive an object in terms of some preconceived notion, in other words to 'project onto' the object characteristics we suppose it to have, even though they are not apparent to the sense organs and may in fact be absent. In so far as this process is integral to all perception, it is normal. Although as a rule the resulting percept is reasonably valid, on some occasions seriously false percepts result.
A second usage is to denote the process whereby a person (male of female) attributes to another (male or female) some features of his own self, especially some aspect of himself that he dislikes or is afraid of. This process must, almost inevitably, lead to false and unfavourable attributions being made about the other person and his motives.
138
There are two reasons for confining the term projection to the second usage. One reason is that another term, 'assimilation', introduced many years ago by Piaget, is already in wide currency to denote our propensity to perceive any object in terms of some model we already have, even though that model may fit the object imperfectly: the new object of perception is said to be assimilated to the existing model. The second reason is that in the various psychoanalytic traditions the most frequent usage of the term projection is to denote our propensity to attribute our failings to others and to be blind to them in ourselves, to see motes in the eyes of others and to be blind to beams in our own.
Using the term in its second sense, we find that the process of projection is invoked extremely frequently by psychoanalysts to explain how it comes about that children and adults should be so afraid, as we know they are, of the wide array of situations that are not intrinsically dangerous. This trend in theorizing has been carried furthest by Melanie Klein who has postulated that the process of attributing to others undesired and frightening features of the self occurs on a major scale during the earliest phases of normal development, with far- reaching effects on later personality. During his first year of life, in the Kleinian
-172-
view, an infant regularly attributes to parent figures impulses that are in fact his own and then introjects (namely creates working models of) parent figures already distorted by these misattributions. In this view, then, the reason a child develops working models of hostile, rejecting, or unresponsive parents ('bad introjected objects') is not so much because of any actual experience he may have had of being unsympathetically or adversely treated by them as principally because, almost from the first, his perception of his parents is gravely distorted by his own prior projections. Since the death instinct is a special aspect of the self that Klein believes is always projected during the earliest months, she is led to a theory of anxiety she sums up in the following sentence: 'I hold that anxiety arises from the operation of the death instinct within the organism, is felt as fear of annihilation (death) and takes the form of fear of persecution' ( Klein 1946). It will be clear that this blanket application of the concept of projection is alien to the present approach. Not only is the Kleinian system of thought rooted in a non-evolutionary paradigm that bears no relation to modern biology, but in clinical work it has the effect, inimical to good practice, of directing attention away from a person's real experiences, past or present, and treating him almost as though he were a closed system little influenced by his environment. Another unfortunate effect of applying the concept of projection in this uncritical way is the danger of bringing a useful concept into disrepute. Let us therefore consider the problem afresh. Not infrequently a person is afraid that someone else intends him harm, but to another's eye this expectation seems misplaced. In such circumstances, as we have seen, psychoanalysts are very apt to postulate that the person who is afraid is projecting onto the other hostile intentions that are in himself but that he denies exist. Though there can be no doubt that this can happen it probably happens much less often than is supposed. In fact a situation of the kind described is explicable in at least four ways; and it is necessary to examine the evidence in each case before deciding which explanation, of which two or more together, is most likely to apply:
1. The subject has rightly detected harmful intent in the other person and in so doing has been more sensitive to the situation than the onlooker.
2. The subject during childhood has learnt that significant
-173-
139
people are often hostile when they claim to be friendly, and is therefore apt, through a process of assimilation, to suppose that figures met with in later life are hostile also when they are not. 3. The subject, aware that he is no friend of the other person and even that he is disposed to do him harm, not unnaturally expects his ill intent to be reciprocated.
4. The subject, unaware of his own ill intent, maintains that, whereas he is friendly to the other, the other is hostile to him.
Of these four possible explanations only the process postulated in the fourth can properly be called projection when the term is used in the restricted sense of attributing to others unwelcome features of the self. That the process can be a source of misattributions is not in doubt. How large a proportion of misattributions have this sort of origin is a matter for inquiry.
The case of Schreber: a re-examination
The urgent need for fresh thinking in this area of psychopathology is shown by the findings of a re-examination by Niederland ( 1959a and b) of the case from which all psychoanalytic theorizing about paranoia and paranoid symptoms derives. Freud's original study of the Schreber case, based solely on the patient's published memoirs, appeared in 1911 ( SE 12: 9- 82). Although he later published other papers on paranoia, according to Strachey ( 1958) Freud never modified his earlier views in any material way.
Daniel Paul Schreber was born in 1842, the second son of an eminent physician and pedagogue. By 1884 he was serving as a judge. He then developed a psychiatric illness from which he recovered after some months. He resumed his legal post but after eight years fell ill again. This time he remained in an asylum for nine years ( 1893-1902) towards the end of which he wrote his memoirs. In 1903, shortly after his discharge, they were published, and soon became a subject of psychiatric interest. A principal theme concerns a number of bodily experiences that were extremely painful and humiliating to him. These experiences he construed as 'miracles' performed by God by means of 'rays':
From the first beginnings of my contact with God up to the present day my body has continuously been the object of divine miracles. . . . Hardly a single limb or organ in my body escaped being temporarily damaged by miracles, nor a
-174-
single muscle being pulled by miracles, Even now the miracles which I experience hourly are still of a nature to frighten every other human being to death. 1
Freud's analysis of Schreber's delusions of persecution takes account of no material except that of the memoirs. Freud notes that Schreber's feelings towards God are intensely ambivalent, being on the one hand critical and rebellious and on the other reverential towards someone of whom he stands in awe. Freud calls attention also to the frankly homosexual attitude Schreber sometimes adopts towards God, including Schreber's belief that he had a duty to play the part of a woman for God's enjoyment. From material of this kind Freud postulates that delusions of persecution are attempts to contradict the proposition 'I (a man) love him (a man)', and to replace it by 'I do not love him -- I hate him', and, finally, by 'I hate him, because he persecutes me'.
140
An internal perception is suppressed, and, instead, its content, after undergoing a certain kind of distortion, enters consciousness in the form of an external perception. In delusions of persecution the distortion consists in a transformation of affect; what should have been felt internally as love is perceived externally as hate.
To this process Freud gives the name projection SE 12: 63-6).
In his re-examination of the case Niederland ( "1959"a and b) draws attention to the fact that Schreber's father held extraordinary views about the physical and moral education of children and published a number of books describing his methods. In them he asserts the vital importance of starting the prescribed re? gimes during infancy and states repeatedly that he has applied his methods to his own children. It is safe, therefore to conclude that Schreber the son had been subjected to his father's educational methods from his earliest years.
The physical methods, recommended for application daily throughout childhood and adolescence, include a number of exercises and harnesses whereby posture is to be controlled. An
____________________
1 In addition to Freud's paper and Strachey's editor's note in the Standard Edition, an English
translation of the memoirs is now available and also a paper by Baumeyer ( 1956) in which he summarizes and quotes from the original case records of Schreber's illnesses. Niederland's bibliography gives references to the above and to the published works of Schreber's father.
-175-
example of a harness, designed to prevent a child's head from falling forwards or sideways, consisted of a strap clamped at one end to the child's hair and at the other to his underwear so that it pulled his hair if he did not hold his head straight. Because the device was apt to produce a stiffening effect it was recommended that its use be restricted to one or two hours a day. An example of an exercise is to place two chairs facing each other with a gap between of a few feet.
A child is instructed to put his head on the seat of one chair and his feet on that of the other and to stiffen his back to make a bridge, in which position he must remain. The dire results that Schreber senior ascribed to bad posture included impeded circulation and, later, paralysis of arm and foot. Of one of his devices, an iron crossbar designed to ensure that a child sits straight, he comments that, besides its physical benefits, it provides an effective moral corrective.
Schreber senior held the sternest of views regarding moral discipline. Bad elements of the mind he regarded as 'weeds' to be 'exterminated', and he describes the threats and punishments by which, starting at five or six months, a parent should make certain that he becomes 'master of the child for ever'. The strong impression given that Schreber senior was a psychotic character is supported by a note made by a hospital psychiatrist and based, it is thought, on information from a member or close acquaintance of the family. It states that the patient's father 'suffered from obsessional ideas with murderous impulses'.
Niederland compares the son's descriptions of the fearful 'miracles' he had to suffer at God's hands with the father's prescriptions of how children should be treated for their physical and moral welfare. Point by point the resemblances are traced. The son complains of miracles of
141
heat and cold. The father prescribes that, in order to toughen an infant, he should be washed in cold water from the age of three months and also subjected to various local cold applications. The son complains that his eyes and eyelids are the target of uninterrupted miracles. The father prescribes repeated visual exercises and advises spraying the eyes with cold water should there be irritation and fatigue following over-stimulation. The son describes a miracle in which his whole chest wall is compressed. Father prescribes a harness consisting of an iron bar that presses against the collar bones should a child not sit straight and upright.
In view of these remarkable resemblances, Niederland's hypothesis, in keeping with the second of the four possible explanations listed above (pp. 173 -4), is that Schreber's delusory
-176-
beliefs regarding the way God was treating him were derived from memories of how his father actually treated him when he was a child. The delusory character of the beliefs is then regarded as due to (a) the patient's attributing the origin of his sufferings to the activities of God in the present instead of to his father in the past, and (b) his attributing the mechanism of his sufferings to 'rays' and miracles instead of to actual manipulations of himself by parent figures. As Niederland ( 1959a) himself remarks, the hypothesis is in keeping with ideas that Freud was entertaining towards the end of his life (but which still have been little exploited). In hallucinations, Freud ( 1937) suggests, 'something that has been experienced in infancy and then forgotten returns. . . . ' .
If this approach to understanding paranoid delusions is adopted, many problems remain still to be solved. How comes it that the patient has no recollection of how his parents treated him as a child? Why is it that, instead, childhood experiences are misplaced in time and the agent responsible for them is misidentified? Possible answers to these questions invoke hypotheses regarding the kinds of injunction, explicit or implicit, a parent may issue to a child; for example, an injunction on a child to construe whatever happens to him as beneficial, an injunction to see his parent as above criticism, an injunction neither to perceive nor to remember certain acts that he none the less witnesses or experiences. These hypotheses, with much evidence to show that they apply to the case of Schreber, are advanced in a recent paper by Schatzman ( 1971). Yet a further hypothesis, not discussed by Schatzman, is that children wish to see their parents in a favourable light and often distort their perceptions accordingly.
In Chapter 20 these matters are pursued further. Meanwhile, enough has been said to show that, when the actual experiences they have had during childhood are known and can be taken into account, the pathological fears of adult patients can often be seen in a radically new light. Paranoid symptoms that had been regarded as autogenous and imaginary are seen to be intelligible, albeit distorted, responses to historical events.
-177-
142
Chapter 12
Fear of Separation
Hypotheses regarding its development
It is now time to draw together ideas regarding fear of separation and how it develops. At the end of Chapter I it is pointed out that 'presence' and 'absence' are relative terms that can give rise to misunderstanding. By presence is meant 'ready accessibility', by absence 'inaccessibility'. The words 'separation' and 'loss' as used in this work imply always that the subject's attachment figure is inaccessible, either temporarily (separation) or permanently (loss). Thus in what follows we are concerned with the developmental processes that lead a young child to respond with fear when he finds, or believes, his attachment figure to be inaccessible. Among the many questions raised and not yet answered are the following:
1. Is inaccessibility of mother in itself a situation that arouses fear in human children without its being necessary for any learning to have taken place?
2. Or is such fear elicited in an individual only after he has come to associate her inaccessibility with a distressing or frightening experience?
3. If the latter, what is the nature of such distressing or frightening experience, and by what type of learning does it become linked with separation?
Whatever the answers to these questions may be, because being alone carries an increased risk of danger, especially for young individuals and others who are weak, the fear response to inaccessibility of mother can usefully be regarded as a basic adaptive response, namely a response that during the course of evolution has become an intrinsic part of man's behavioural repertoire because of its contribution to species survival.
If that is so, there is no a priori reason to assume that fear elicited by mother's inaccessibility can be explained only in terms of an individual's having experienced something dis-
-178-
tressing or frightening while separated from her, an assumption that has commonly been made. On the contrary, it is entirely possible that the response to mother's inaccessibility develops during ontogeny without learning of any sort having to take place. Let us call this hypothesis A.
Whether or not hypothesis A applies in the human case remains an open question. For, as was emphasized repeatedly in the first volume, there are many forms of behaviour that, like this one, can usefully be classed as instinctive but that develop functionally only when the environment provides opportunity for learning of some specific kind to occur. In other words, to hold the hypothesis that fear behaviour in a situation of maternal inaccessibility is instinctive in no way rules out the possibility that learning of some kind is necessary for its development. All that such a view requires is that, when an individual is reared in the species's environment of evolutionary adaptedness, opportunity for the necessary learning is always present.
143
Reflection suggests that there are at least three hypotheses that are consistent with that proviso and merit attention. Let us call them hypotheses B1, B2, and B3.
The first, B1, is Freud's hypothesis of 1926 which postulates that fear of mother's absence results from an infant's learning that, when she is absent, his physiological needs go unmet, and learning, further, that this results in the accumulation within him of dangerous 'amounts of stimulation' which, unless 'disposed of', bring about a 'traumatic situation'. Since, moreover, the infant finds that, left to himself, he is unable to dispose of such accumulations, the danger situation that he comes intrinsically to fear is 'a recognized, remembered, expected situation of helplessness'. 1
Reasons for not adopting Freud's hypothesis will already be apparent. One is that it is embedded in a paradigm very different from the one adopted here (see Chapter 5). Another is that it seems to postulate a degree of insight into cause and effect that not only is improbable in an infant of a year or so of age but that we now know to be unnecessary to account for the findings. For the fact that so many of the responses shown by a human infant when separated from his mother are to be seen also in infants of non-human primate species demonstrates that
____________________
1 Quotations are from Inhibitions, Symptoms and Anxiety ( SE 20: 137-8 and 166). Freud's
theory is described more fully in Appendix I. -179-
it is quite possible for such responses to be mediated at a primitive and presumably infra- symbolic level.
Objections of a similar sort apply to the theories advanced by Klein, which presuppose even more sophisticated cognitive functioning (see Appendix I).
A second hypothesis, B2, not very different from Freud's but simpler and implying no insight learning, is compatible with the theory of attachment behaviour proposed in the first volume. In Chapter 14 of that volume an account is given of the conditions that terminate crying during the early months of life:
. . . when a baby is not hungry, cold, or in pain, the most effective terminators of crying are, in ascending order, sound of voice, non-nutritive sucking, and rocking. These findings readily explain why babies are said to cry from loneliness and to have a desire to be picked up. Although to attribute such sentiments to babies in the early months of life is almost certainly not warranted, the statements none the less contain more than a grain of truth. When they are not rocked and not spoken to infants are apt to cry; when they are rocked and spoken to they cease crying and are content. And by far the most probable agent to rock and talk to a baby is his mother figure.
In view of this, it could be argued, an infant comes to learn that presence of mother is associated with comfort while absence of mother is associated with distress. Thus, through a fairly simple process of associative learning, an infant comes to associate mother's absence with distress, and so to fear her being inaccessible. This hypothesis is close to one advanced by Kessen & Mandler ( 1961).
144
A third hypothesis, B3, derives from the fact that an infant is much more intensely afraid of fear-arousing situations, such as strangeness, or sudden approach, or loud noise, when his mother is absent than when she is present. After a few such experiences, it could be postulated, mother's absence might of itself come to elicit fear, again through a process of associative learning. This hypothesis is similar to one suggested by Rycroft ( 1968a) and referred to in Chapter 6 above.
On present evidence it is not possible to decide between hypotheses A, B2, and B3; each is plausible.
Hypothesis A, that an anxiety response to mother's inaccessibility develops during ontogeny without learning of any
-180-
sort having to take place, is difficult to test. Furthermore, even if it were true, it would not make hypotheses B2 and B3 irrelevant, since learning of the kinds proposed by these two hypotheses could still occur and might be of much significance in accounting for degrees of separation anxiety above a minimum.
Whether hypothesis A is valid or not appears, at present, to be of no great clinical importance. This is because, were the forms of learning postulated by hypotheses B2 and B3 to occur at all, which they probably do, they would be taking place during the latter half of the first and during the second year of life and, except where a child had no mother figure, would be virtually unavoidable. As a situation that arouses fear, therefore, separation from an attachment figure would still be nearly universal, almost as much as it would were hypothesis A to apply.
Support for the view that associative learning of the kinds postulated by hypotheses B2 and B3 does take place comes from studies of individual differences in susceptibility to respond with fear, especially to separation. These show, as is discussed in detail in later chapters, that children who have been well mothered, and therefore, in all likelihood, have been protected from the experience both of intense distress and of intense fear, are those least susceptible to respond with fear to situations of all kinds, including separation; whereas children who have had intensely distressing and frightening experiences when away from mother are apt to show an increased susceptibility to fear, especially to fear of being separated again.
Should, as therefore seems likely, both these forms of associative learning occur during infancy and early childhood, their effects on personality development might possibly be rather different. For example, were a child, because of his particular experiences, to come to associate mother's absence with high degrees of discomfort and distress, he might perhaps grow up to respond to separation and loss, either actual or forecast, with psychosomatic troubles and general tension; whereas a child who, because of his particular experiences, came to associate mother's absence with being more or less intensely afraid might grow up prone to respond to any fear-arousing situation with more marked fear than would other individuals. Whether or not differences of these kinds occur in fact can be determined only by further research.
-181-
145
Need for two terminologies
Throughout recent chapters a sharp distinction has been drawn between situations that arouse fear and situations that are intrinsically dangerous. Whereas situations that arouse fear can be regarded as constituting either natural or cultural clues to an increased risk of danger, they are certainly not infallible indicators of actual danger. How we feel in a situation bears therefore only an indirect relationship to the degree of risk present in that situation. Because the world as reflected in feeling is distinct from, though correlated with, the world as it is, two terminologies are necessary. At the end of Chapter 6 three terms, 'anxious', 'alarmed', and 'afraid' are introduced and the way in which they are used here is described. All three belong to the world as reflected in feeling. By contrast, 'dangerous' belongs to the world as it is. At this point it is necessary to settle on some analogously distinct terms suitable to refer, on the one hand, to a state of feeling antithetical to feeling afraid and, on the other, to a situation antithetical to one of danger. Etymology suggests 'feeling secure' for the one and a 'situation of safety' for the other. The original meaning of the English adjective 'secure' is 'free from care, apprehension, anxiety or alarm' ( Oxford English Dictionary). Historically, therefore, 'secure' applies to the world as reflected in feeling and not to the world as it is. By contrast, the original meaning of 'safe' is 'free from hurt or damage'. As such it applies to the world as it is and not to the world as reflected in feeling. The distinction is neatly illustrated by a seventeenth-century saying, quoted in the OED, 'The way to be safe is never to bee [sic] secure', namely feel secure. By using the terms in their original senses, it is possible accurately and without ambiguity to make statements such as:
--although the situation was safe enough he became very frightened, or
--I could see the situation was dangerous but somehow the captain's behaviour made us all feel secure.
The distinction drawn here between feeling secure and being safe is not always made so that a number of terms current in the literature do not conform to the usage proposed. This applies both to Harlow's 'haven of safety', termed here 'secure base',
-182-
and to Sandler's 'feeling of safety' ( Sandler 1960), termed here 'feeling of security'.
Use of the word 'secure' in the sense proposed has, of course, for long been customary in clinical practice. For example, with reference to states of feeling, children and grown-ups are habitually described as being either secure or insecure. Moreover, because any person who is acting as an attachment figure for another is commonly referred to as providing that other with a sense of security, it is often convenient to describe an attachment figure also as a security figure or as providing a secure base. At the same time, it must be emphasized that a secure base, however much it may lead someone to feel secure, is no guarantee of safety, any more than a natural clue, however frightening we find it, is a certain indicator of danger. As a guide to what is safe and what is dangerous the kind of feeling a situation arouses in us is never more than rough and ready.
-183-
146
Part III
INDIVIDUAL DIFFERENCES IN SUSCEPTIBILITY TO FEAR: ANXIOUS ATTACHMENT
-185-
Chapter 13
Some Variables responsible for Individual Differences Constitutional variables
That individuals differ enormously in their susceptibility to respond to situations with fear is a commonplace. Why they should differ in such extreme ways remains a puzzle. In this chapter and those following an attempt is made to identify some of the many variables that are operating. The main focus, of course, is on the part played by a person's relationship to his attachment figure(s). This is held to be pervasive and still too little understood. Let us consider first some of the other variables. It must be assumed that genetic differences play some part in accounting for variance between individuals with regard to susceptibility to fear. Very little is yet known about their role in humans, but it is well documented in the case of other mammals, e. g. dogs ( Scott & Fuller 1965; Murphree, Dykman & Peters 1967). A difference in susceptibility in humans that is likely to be in part genetically determined is one between men and women.
Sex Differences
Feminist opinion notwithstanding, it is very commonly believed that there are some differences in susceptibility to fear as between men and women. This view is plausible and there is some evidence to support it. At the same time it is clear that in this regard there is much overlap between any population of women and a comparable population of men. Culture, moreover, can either magnify such potential differences as there may be, for example by sanctioning the expression of fear by members of one sex but not by those of the other, or else try to reduce them. Evidence from four sources supports the idea of a difference in susceptibility between the sexes:
In the experiments with nursery-school children, carried out by Jersild & Holmes ( 1935a) and described in Chapter 7,
-187-
a higher percentage of the girls were afraid than of the boys. The situations in which the difference was most marked were going into the dark passage and approaching the two animals, snake and dog. In these three situations the percentages of boys who showed fear were respectively 36, 40, and 46. The comparable percentages for girls were 48, 50, and 59.
In interviews of mothers of children aged six to twelve years Lapouse & Monk ( 1959) found that the proportion of girls reported as being afraid of strangers and animals, notably snakes, was higher than that of boys. In two other studies in which children of about the same age were interviewed, girls reported more situations as feared than did boys ( Jersild, Markey & Jersild 1933; Croake 1969).
147
In questionnaires given to students there is a consistent tendency for women to report more situations as feared than men (for references and comment see Marks 1969).
In epidemiological studies of psychiatric casualties women are reported to suffer from anxiety states about twice as frequently as men ( Leightonet al. 1963; Hare & Shaw 1965). Two-thirds of agoraphobic patients seen by psychiatrists are women ( Marks 1969).
A difference in the opposite direction -- that females tend to show less fear than do males -- seems not to have been reported.
Viewed in an evolutionary perspective these findings are not surprising. In most races of man, as in other species of groundliving primates, males are larger and stronger than females ( Cole 1963). While males bear the brunt of defence against predators, as well as attacking them when necessary, females protect young and, unless prevented from doing so, are more likely to retire from dangerous situations than to grapple with them. It would be strange were such long-standing differences between the sexes in respect of body structure and social role not to be reflected in complementary differences in behavioural bias.
Minimal Brain Damage
In Chapter 16 of the first volume an account is given of a longitudinal study of twenty-nine pairs of boys ( Ucko 1965), which shows that children who at birth are noted to be suffering from asphyxia are much more sensitive to environmental change than are matched controls. When the family went on holiday
-188-
or changed house, boys who had suffered from asphyxia were more likely to be upset than were the controls. The same was true when a member of the family -- father, mother, or sibling -- was absent for a time. These differences were apparent during each of the first three years of life (though not significantly so during the third). A comparable difference was seen when some of the children started nursery school.
Soon after his fifth birthday every child started infant school, making this the only event that was common to them all (though of course they went to many different schools). Here again the difference between the two groups was striking and significant. On a three-point scale (reduced from five points), the children distribute as shown below:
Enjoyed start accepted it
Mild
protest
within one week
Mild
school from the
or
at least
apprehension and disappearing
8 10 apprehension or 13 2
Asphyxiated at birth
Controls
8 17
148
marked
more than a week
Totals
Childhood Autism
disturbance
lasting
The behaviour of an autistic child shows a complete absence of attachment together with many indications of chronic fear. Tinbergen & Tinbergen ( 1972), adopting an ethological approach, suggest that the underlying condition may be one of chronic and pervasive fear, which cannot be allayed by contact with an attachment figure because the child also fears humans. If this is so, the syndrome could be conceived as resulting from a persistently lowered threshold to fear-arousing stimuli combined with delayed development of and/or inhibition of attachment. Causal factors might then include any of
-189-
the following: (a) genetic factors, (b) brain damage, (c) inappropriate mothering. A combination of two or more factors seems likely. Clancy & McBride ( 1969) describe a treatment programme based on this type of theory.
Blindness
Nagera & Colonna ( 1965) report that blind children are apt to be more than usually afraid of such common fear-arousing situations as animals, mechanical noises, thunder and wind, and to live in a state of permanent alertness. A principal reason for this is probably that, being blind, they are likely to be out of contact with their attachment figure far more often than are sighted children, and thus often to be effectively alone when something frightening occurs. Their tendencies on some occasions to remain rigidly immobile and, on others, to seek very close bodily contact with an adult are in keeping with this explanation.
Great difficulties arise for such children after a brief separation because a blind child cannot track his mother visually and keep close to her as a sighted child commonly does on such occasions. Fraiberg ( 1971) describes the very acute reaction of a blind boy of fourteen months after his mother had been absent for three days, during which he had been cared for by various friends and relations. During the first fortnight after mother's return he screamed for hours at the highest pitch, 'something between terror and rage', or else shouted and chanted perpetually. Only when his mother held him was there any respite; and then he would crawl relentlessly all over her. Because the screaming was so distressing to mother it was suggested she give him pots and pans to bang together instead. This the child did with great gusto and the screaming ceased.
Fraiberg describes also another blind child, a little older, who was cared for by familiar grandparents while mother had a new baby. When reunited with his mother he was markedly ambivalent at first but responded quickly when she, an affectionate mother, gave him plenty of cuddling. The main reason for the far more acute reaction in the younger child is likely to have been that his mother was a disturbed woman whose mothering was erratic both before
Asphyxiated at birth
Controls
29 29
149
and after her absence; another factor may have been that he was cared for by several different people while she was away.
-190-
Changes during Development in a Child's Susceptibility to Fear
While every infant comes into the world with biases to respond in some ways more than in others, how he develops turns on a process of interaction between himself and his environment. In regard to a susceptibility to respond fearfully, there are certain developmental trends sufficiently buffered to environmental variation to be seen in a huge majority of individuals. For example, as related in Chapter 7, all descriptive studies agree that, whereas during the first two years of life a child is broadening the range of situations he fears -- to include especially strangeness, animals, darkness, and separation -- from his fifth birthday onwards, and often before, he is likely to become steadily more discriminating in what he fears and more confident and competent in dealing with situations that would formerly have frightened him. Because change towards greater discrimination and confidence represents the norm, we start by considering the nature of the experiences and processes likely to be responsible for it. Subsequently we consider experiences and processes that have an opposite effect, for example, those that interfere with the usual tendency for susceptibility to diminish, or even enhance the susceptibility, and others that have the effect of increasing the range of situations feared.
Experiences and processes that reduce susceptibility to fear
The experiences that occur and the processes at work during the ordinary course of a person's life that tend to reduce his susceptibility to fear are of many kinds. A principal process, increasing confidence in the availability of his attachment figure(s), is the subject of the next chapter. Of the others the main ones can be described, in everyday language, as getting used to situations that are initially alarming, discovering that in many such situations other people are not afraid, and learning to tackle a situation actively and thereby discovering that nothing ill befalls. In the language of learning theory they are termed:
-- habituation
-- observational learning leading to vicarious extinction
-- observational learning combined with guided participation.
Other processes are likely to be at work as well, though it is not clear how large a part they play during the ordinary course
-191-
of development. For example, it may well be that there is some naturally occurring version of the procedure developed by behaviour therapists, and known variously as 'reciprocal inhibition', 'counter-conditioning', and 'desensitization', in which an association is gradually built up between a stimulus situation that is feared and something that the subject finds pleasant. 1
Yet another process, and one that it is easy to forget, is that as an individual grows up he becomes stronger and more skilful so that situations that might once have been, or at least seemed, dangerous to him cease to be so.
Knowledge of some of these processes has been greatly extended in recent years by the work of learning theorists and behaviour therapists. As Marks ( 1969) is at pains to stress, a large
150
majority of these studies have been conducted with healthy individuals who happen to be intensely afraid of some delimited object or situation, such as a snake or a dog, and not with psychiatric patients who commonly suffer not only more generalized anxiety but usually difficulties in personal relationships and a tendency to depression as well. It is for this reason that many clinicians suspect that the findings of the learning theorists may prove to be of only limited value in psychiatric practice. Yet it is for this same reason that their findings are in all likelihood of much relevance in understanding how it happens that the tendency to respond fearfully recedes during the ordinary course of healthy development.
Let us consider further the three processes already listed.
Habituation
This is a process of learning not to respond to a situation when it is followed by nothing of consequence. It presumably plays a major part in restricting an infant's initial tendency to respond with fear to all and any strong or sudden stimulation. Later, habituation, perhaps in more sophisticated forms, also restricts the range of situations that are responded to with fear because they are strange; for much of what is strange today will not only have become familiar tomorrow but also have been found to lead to no untoward consequence. Thus habituation greatly limits the range of situations responded to with fear. It should none the less be noted that habituation in no way affects
____________________
1 A full description of desensitization and related techniques is given in Marks ( 1969).
-192-
the basic and persistent tendency to respond with fear, as well as with curiosity, to anything perceived as strange.
Observational Learning leading to Vicarious Extinction
It has already been remarked that observational learning can work in either of two directions: either the observer learns to fear situations that formerly he did not fear or else he learns not to fear situations that formerly he did fear. The most important component in learning not to fear situations formerly feared, Bandura ( 1968) finds, is that the observer should see that the feared situation can be approached and dealt with without there being any bad consequences. The identity of the person observed (model) and the degree to which the observer can identify with him are found to be of much less significance. Even watching a sequence on film can have a reassuring effect, provided always that the consequences of the model's actions are clearly depicted.
The process of learning that something is harmless from direct observation of the experience of others is very different, it should be noted, from merely being informed by another person that a situation is harmless. All those who have made a systematic study of the problem report that simple explanation and reassurance have only very limited effect, a finding that will come as no surprise to clinicians.
Fortunately, in the ordinary course of events, a child growing up in a family has endless opportunities to learn from observation that many of the situations that make him afraid are in
151
fact harmless. Parents, older brothers and sisters, neighbours and schoolfellows are continually and without knowing it providing a child with this indispensable information.
Observational Learning combined with Guided Participation
This method requires much more from the model than giving the subject opportunity for simple observational learning. It is evident none the less that every sensible parent is constantly providing it. The method consists in the model's first demonstrating in action that the feared situation holds no danger and then encouraging the other person -- child or adult -- to tackle the situation himself. Once again it appears that the crucial part of the process is that the learner should discover, this time for himself, that approaching and tackling the situation can be done without untoward consequences.
--sight of opera glasses and sound of whistle together were not alarming, nor was there reason to think that either singly would be
--sight of the caterpillar aroused interest which alternated with slight fear
--sight of caterpillar and sound of whistle together were alarming.
From observations such as these Valentine suggests that there is a much readier bias to develop fear of objects like caterpillars than fear of objects like opera glasses. 1
The extreme ease with which monkeys and apes develop a fear of snakes is remarked on above (Chapter 8, p. 130). The same is true of humans. As already reported in Chapter 7 (pp. 109-13), in the experiments conducted by Jersild & Holmes ( 1935a) between one-third and a half of the children aged between two and six years showed marked fear of the snake. A comparable finding is reported by Morris & Morris ( 1965).
____________________
1 A defect of this experiment was that during the second test when the caterpillar was
presented and the whistle blown Y was seated on her father's knee, not her mother's. It is therefore possible that a change in the person caring for her was responsible for the results.
-163-
132
In a children's television programme in England children were invited to compete for a prize by proposing future programmes. To qualify, however, they had to name the animal they liked most and the animal they disliked most. Altogether nearly twelve thousand children aged from four years upwards replied. Of the animals most disliked the snake was an easy first: it was named by 27 per cent of the children. Spiders came next, given by under 10 per cent; then lions and tigers, together, were named by about 7 per cent. Up to the age of nine years, at least one child in three expressed dislike of snakes. At all ages slightly more girls than boys expressed the fear.
It seems likely that to the development of a fear of animals in general and of snakes in particular several factors make an interlocking contribution. First are several of the common natural clues, including often strangeness. Second, there may also be certain specific natural clues, for example crawling or wriggling. Third, there is the behaviour of others. Because of their appearance and behaviour, including their vocalizations, animals arouse simultaneously both lively interest and incipient fear. In such conditions the behaviour of a companion is likely to have a maximum effect, tipping the balance either towards decreased fear and approach or towards increased fear and withdrawal.
Fear of Darkness
Every study shows that fear of darkness is as common at every age as fear of animals and that during ontogeny it runs a roughly parallel course. In all likelihood the development of fear of darkness is to be explained in a comparable way to that of fear of animals; though the natural clues concerned are usually not the same ones.
In conditions of darkness the two natural clues that are apt to be present together are strangeness and being alone. During darkness visual stimuli that, if viewed in daylight, would be recognized as familiar are often ambiguous and difficult to interpret. Endless examples come to mind: the pattern of movement of light shining through bedroom curtains; the shapes of the trees in a wood at night; the shadowy recesses of a dimly lit cellar. In each case the visual stimuli available are barely adequate for accurate perception and it is therefore as easy to perceive something unusual as something familiar. In addition, without visual cues, sounds are far more difficult to interpret accurately or with confidence. Thus in conditions of
-164-
darkness much seems uncertain or strange and in consequence alarming.
Yet, mere strangeness would probably of itself arouse comparatively little fear were it not so regularly accompanied by being alone. Sometimes a person is in reality alone; sometimes, because his companion is unseen, he may merely feel himself to be. In either case the situation is compound: it combines sights and sounds that are not easily interpreted and the situation of being alone.
Freud, it is interesting to note, was greatly struck by the way in which darkness leads a child to feel alone, and it is an observation of a small boy's behaviour in the dark, together with inferences from it, that lies at the heart of his theory of anxiety. This is therefore a good point at which to compare Freud's theory with the one presented here.
133
In Three Essays ( 1905b, SE 7: 224n) and again in the Introductory Lectures ( 1917b, SE 16: 407) Freud tells the story of a three-year-old boy. He recounts how he once heard this child
calling out of a dark room: 'Auntie, speak to me! I'm frightened because it's so dark. ' His aunt answered him: 'What good would that do? You can't see me. ''That doesn't matter,' replied the child, 'if anyone speaks, it gets light. ' Thus [comments Freud] what he was afraid of was not the dark, but the absence of someone he loved . . .
Reflection on that episode, Freudtells us, led him to the view that the prototypic situation that gives rise to anxiety in children is simply separation from mother. Neurotic anxiety, he then argues, can best be understood as a persistence beyond childhood of the tendency to be anxious when alone, though fear of being alone often masquerades as fear of something else, for example of the dark. In all these regards the theory advanced here is very close to Freud's. Where the two differ is that Freud did not recognize that strangeness is intrinsically frightening or that both strangeness and being alone can usefully be regarded as two members of a class of natural clues to increased risk of danger. As a consequence he held that to be afraid when alone (and also when confronted by any of the other natural clues) is irrational and neurotic; whereas in the theory put forward here to be afraid in such conditions is held to be in general adaptive.
-165-
Fear of Being Alone
Repeatedly in these chapters it is stressed that being alone is one of several natural clues to increased risk of danger and that it occurs very commonly as a component in a compound situation. Not only does it occur in combination with other natural clues, moreover; it can occur equally well with cultural clues and also in situations realistically assessed as being dangerous. Thus, throughout life, being alone is a condition that either stimulates fear or greatly intensifies fear aroused in other ways. Concomitantly, being with a companion greatly reduces fear. In no conditions is the reassuring effect of the presence of companions more evident than during and after a disaster.
Behaviour in disaster
The role of a companion in reducing fear in children is very obvious and it is also readily acknowledged by children. Adults, by contrast, are less likely to acknowledge it. During and after a disaster, however, people are less reticent ( Baker & Chapman 1962).
When the impact of disaster comes members of a family commonly cling together:
When sirens scream of approaching disaster, minds turn to loved ones. If they are near enough mothers run to protect their children, and men seek their families. They huddle together and support one another through the stress, and when it is passed they resume and nurse those they love ( Hill & Hansen 1962).
Wolfenstein ( 1957) describes how a woman who had been with her fifteen-year-old daughter when a tornado struck recounted her experience:
134
And she said, 'Mother, its coming -- a cyclone. ' And I said, 'Mary, I'm afraid it is. But,' I said, 'we're together. ' And she said, 'Mother, I love you and we're together. ' I shall never forget those words. And we -- our arms were around each other, and I said, 'Whatever happens, Mary, let's cling together. '
Whenever members of a family happen to be apart at the time of impact they are unlikely to rest until they have found each other; then, again, physical embraces are the rule. 'Just
-166-
being together is deeply important following impact even in loosely-knit families' ( Hill & Hansen 1962).
Survivors agree that to be alone during a disaster is extremely frightening, whereas the advent of a companion, however inadequate, is likely to transform the scene. Wolfenstein refers to another episode, one in which, following an explosion, two injured men were trying to crawl out of a burning factory. Describing their experiences, one of the men, who had suffered a broken leg, explained:
Then Johnny and Clyde came along. I said, 'Johnny, help us -- we can't walk. ' His arms were broken and he said, 'I can't help you, but I'll stay with you. If you can crawl, I'll guide you. ' Talk about cheer! That helped me more than anything -- just when he said, 'I'll stay with you. '
Not only is there a strong tendency for members of a family or other social group to remain together during the height of a disaster but the tendency is likely to persist for days or weeks after it is over. This heightened tendency to attachment behaviour is commented on in a number of reports.
For example, Bloch, Silber & Perry ( 1956) studied the effects on children of a tornado that struck a town in Mississippi, affecting in particular a cinema in which children were attending a Saturday afternoon programme. Altogether, information was obtained during the succeeding weeks from interviewing the parents of 185 children between the ages of two and twelve years.
About one-third of the children were reported to be showing signs of increased anxiety, which typically took the form of clinging to or remaining close to parents and wishing to sleep with them. They were made anxious by noise and also tended to avoid situations associated with the tornado. Children aged six to twelve were more disturbed than younger ones. A possible reason for that was that more of them had probably been in the impact zone. Another possible reason, though it is not commented on by the researchers, was that the older the child the more likely was he to have been away from his parents. Boys were as much affected as girls.
Experiences that were significantly associated with increased anxiety were the child's presence in the impact zone, personal injury, and the death or injury of a family member. Not unexpectedly the reaction of children reflected the reaction of parents. In nine cases parents described themselves as having
-167-
135
'gone to pieces' and, instead of having supported their child, as having sought help from him. Eight of these children were disturbed and, about the ninth, the mother could not be induced to talk. Further discussion of the contribution to a child's anxieties of parents who invert the relationship by requiring their child to care for them will be found in Chapters 18 and 19. Many cases diagnosed as school phobia and agoraphobia can be understood as being caused by such inversions.
Reports on the effects of the Mississippi tornado of 1953 ( Blochet al. 1956) and of the Los Angeles earthquake of 1971 ( Time, 8 March 1971) both make it clear that, after a disaster, parents are almost as eager to retain their children close to them as the children are eager to remain close to their parents. Since these responses are adaptive, it is unfortunate that the concept of regression is so frequently invoked to explain them. Investigation shows that, in run-of-the-mill cases as well as in those that follow a disaster, behind behaviour dubbed regressive by clinicians there exist situations that, once known about, explain at once why a child or an adult should cling relentlessly to another member of his family.
-168-
Chapter II
Rationalization, Misattribution, and Projection
All round the house is the jet-black night;
It stares through the window-pane;
It crawls in the corners, hiding from the light, And it moves with the moving flame.
Now my little heart goes a-beating like a drum,
With the breath of the Bogie in my hair;
And all round the candle the crooked shadows come, And go marching along up the stair.
The shadow of the balusters, the shadow of the lamp, The shadow of the child that goes to bed --
All the wicked shadows coming, tramp, tramp, tramp, With the black night overhead.
ROBERT LOUIS STEVENSON, A Child's Garden of Verses
Difficulties in identifying situations that arouse fear
When a person is afraid and claims that something in particular, for example thunder or a dog, has made him so, doubt is often expressed whether he has identified the right stimulus situation. This is especially likely to happen when fear is shown or reported by children and by emotionally disturbed adults. Among psychoanalysts there is a long tradition of claiming that what a person is really afraid of is something very different from what he claims to be afraid of. Indeed, psychoanalytic theorizing about anxiety and fear reflects a prolonged hunt for some primal danger situation that is thought to arouse a primal anxiety or fear. 1 Arising out of that tradition, also, is the practice of invoking the process of projection whenever a fear appears not to be appropriate to the situation presenting.
136
In the approach adopted here, no less than in the traditional ones, misattribution is held to be very common. The difference in the present approach lies in the explanations it offers of why misattribution should occur. The concept of a primal danger
____________________
1 See Chapter 5 and Appendix I.
-169-
situation is dispensed with, and projection is given a much smaller role as an explanatory principle. A solution is found in the relationship that the natural clues bear to danger and safety.
The very fact that fear is first aroused in human beings not by any rational appraisal of danger but by stimulus situations that are no more than clues to an increased risk of danger invites misunderstanding and misattribution; for, as has been made clear, a natural clue is in no sense inherently dangerous. Because that is so, however, and because in Western culture (and perhaps also in others) a human being is expected to be afraid only of real dangers, there is a strong bias both in the frightened subject and in an onlooker to attribute the fear response to something other than the natural clue. For example, since it is thought absurd for anyone to be afraid merely of thunder, the fear is 'explained' as a fear really of being struck by lightning. Similarly, since it is thought absurd to be afraid merely of a dog, the fear is 'explained' as a fear really of being bitten by a dog.
Rationalizations of these sorts are no doubt very common. They are commented upon by all who have made a study of fear, irrespective of theoretical orientation. For example, Marks ( 1969) suggests that a child's fear of monsters in the dark may be no more than a rationalization of his fear of the dark, 'a genuine rationalization of an irrational fear on the same lines that any post-hypnotic suggestion is rationalized'. The Newsons ( 1968) point out that such rationalizations are often and easily encouraged by other children, or even by adults, who tease a child about what he might meet when in the dark on his own. Jersild ( 1943) calls attention to the fact that when a child is already frightened, from whatever cause, he 'may formulate his fear in terms of an imaginary or anticipated danger', such as criminal characters or bogeys or some other sinister circumstances he has encountered or, more probably, heard or read about.
Though simple rationalizations of this kind are probably common, even commoner perhaps are mistaken or biased attributions stemming from the special properties of compound situations. In compound situations, two or more stimulus conditions, by being present together, have the effect of arousing far more intense fear than would any one were it to occur separately. In such a case there is a marked tendency to single out one component of the compound situation as the one that arouses the fear and to ignore the other(s). For example, a
-170-
person is alone in the dark and hears strange noises. Whereas in fact all three conditions -- being alone, being in the dark, and hearing the strange noises -- may well be necessary to account for the fear aroused, in all likelihood attention is focused solely on the strange noises and the other components of the situation are almost ignored. Furthermore, from there it is only a short step to rationalize the fear, aroused in fact by what is little more than a
137
combination of two or three natural clues, and to claim that what is feared is burglars or ghosts.
Which one of several components present in a compound situation is fastened on as the fear- arousing one and which are neglected needs examination. Presumably, the component selected is usually the one that most readily lends itself to being interpreted as indicative of real danger. If that is so, being alone would habitually be neglected, or at least given a subordinate position. That is, in fact, very close to what Freud believed to happen, though he expresses his views in terms of libido theory and not of attachment theory.
In 1917, at the end of a discussion of the psychopathology of phobias, Freud summed up his position:
Infantile anxiety has very little to do with realistic anxiety, but, on the other hand, is closely related to the neurotic anxiety of adults. Like the latter, it is derived from unemployed libido, and it replaces the missing love-object by an external object or by a situation ( 1917b, SE 16: 408)
Since Freud regards the unemployed libido as constituting an internal danger, his formulation is that fear of an internal danger is replaced by fear of an external one. An alternative rendering of his position would run: when a child or an adult is afraid of some external object or situation, what he is really afraid of is the absence of someone he loves.
In Chapters 18 and 19, in which misattributions are discussed further, reasons are given for believing that many intense fears attributed to all sorts of common situations and termed phobias are best understood as being aroused in compound situations, a main component of which is the expectation of being separated from a principal attachment figure. The famous and theoretically influential case of 'Little Hans', who was afraid of being bitten by a horse, appears to be a good example ( Freud 1909, SE 10). Evidence is presented (Chapter 18) for believing that fear of separation played a much larger part in this case than Freud at the time realized.
-171-
Misattribution and the role of projection
In some psychoanalytic traditions the concept of projection has been used very extensively in an attempt to explain any fear that is not readily intelligible as a response to a real danger. Since the term is itself used in several ways the resulting theory is often confused.
One usage of projection is to denote our propensity to perceive an object in terms of some preconceived notion, in other words to 'project onto' the object characteristics we suppose it to have, even though they are not apparent to the sense organs and may in fact be absent. In so far as this process is integral to all perception, it is normal. Although as a rule the resulting percept is reasonably valid, on some occasions seriously false percepts result.
A second usage is to denote the process whereby a person (male of female) attributes to another (male or female) some features of his own self, especially some aspect of himself that he dislikes or is afraid of. This process must, almost inevitably, lead to false and unfavourable attributions being made about the other person and his motives.
138
There are two reasons for confining the term projection to the second usage. One reason is that another term, 'assimilation', introduced many years ago by Piaget, is already in wide currency to denote our propensity to perceive any object in terms of some model we already have, even though that model may fit the object imperfectly: the new object of perception is said to be assimilated to the existing model. The second reason is that in the various psychoanalytic traditions the most frequent usage of the term projection is to denote our propensity to attribute our failings to others and to be blind to them in ourselves, to see motes in the eyes of others and to be blind to beams in our own.
Using the term in its second sense, we find that the process of projection is invoked extremely frequently by psychoanalysts to explain how it comes about that children and adults should be so afraid, as we know they are, of the wide array of situations that are not intrinsically dangerous. This trend in theorizing has been carried furthest by Melanie Klein who has postulated that the process of attributing to others undesired and frightening features of the self occurs on a major scale during the earliest phases of normal development, with far- reaching effects on later personality. During his first year of life, in the Kleinian
-172-
view, an infant regularly attributes to parent figures impulses that are in fact his own and then introjects (namely creates working models of) parent figures already distorted by these misattributions. In this view, then, the reason a child develops working models of hostile, rejecting, or unresponsive parents ('bad introjected objects') is not so much because of any actual experience he may have had of being unsympathetically or adversely treated by them as principally because, almost from the first, his perception of his parents is gravely distorted by his own prior projections. Since the death instinct is a special aspect of the self that Klein believes is always projected during the earliest months, she is led to a theory of anxiety she sums up in the following sentence: 'I hold that anxiety arises from the operation of the death instinct within the organism, is felt as fear of annihilation (death) and takes the form of fear of persecution' ( Klein 1946). It will be clear that this blanket application of the concept of projection is alien to the present approach. Not only is the Kleinian system of thought rooted in a non-evolutionary paradigm that bears no relation to modern biology, but in clinical work it has the effect, inimical to good practice, of directing attention away from a person's real experiences, past or present, and treating him almost as though he were a closed system little influenced by his environment. Another unfortunate effect of applying the concept of projection in this uncritical way is the danger of bringing a useful concept into disrepute. Let us therefore consider the problem afresh. Not infrequently a person is afraid that someone else intends him harm, but to another's eye this expectation seems misplaced. In such circumstances, as we have seen, psychoanalysts are very apt to postulate that the person who is afraid is projecting onto the other hostile intentions that are in himself but that he denies exist. Though there can be no doubt that this can happen it probably happens much less often than is supposed. In fact a situation of the kind described is explicable in at least four ways; and it is necessary to examine the evidence in each case before deciding which explanation, of which two or more together, is most likely to apply:
1. The subject has rightly detected harmful intent in the other person and in so doing has been more sensitive to the situation than the onlooker.
2. The subject during childhood has learnt that significant
-173-
139
people are often hostile when they claim to be friendly, and is therefore apt, through a process of assimilation, to suppose that figures met with in later life are hostile also when they are not. 3. The subject, aware that he is no friend of the other person and even that he is disposed to do him harm, not unnaturally expects his ill intent to be reciprocated.
4. The subject, unaware of his own ill intent, maintains that, whereas he is friendly to the other, the other is hostile to him.
Of these four possible explanations only the process postulated in the fourth can properly be called projection when the term is used in the restricted sense of attributing to others unwelcome features of the self. That the process can be a source of misattributions is not in doubt. How large a proportion of misattributions have this sort of origin is a matter for inquiry.
The case of Schreber: a re-examination
The urgent need for fresh thinking in this area of psychopathology is shown by the findings of a re-examination by Niederland ( 1959a and b) of the case from which all psychoanalytic theorizing about paranoia and paranoid symptoms derives. Freud's original study of the Schreber case, based solely on the patient's published memoirs, appeared in 1911 ( SE 12: 9- 82). Although he later published other papers on paranoia, according to Strachey ( 1958) Freud never modified his earlier views in any material way.
Daniel Paul Schreber was born in 1842, the second son of an eminent physician and pedagogue. By 1884 he was serving as a judge. He then developed a psychiatric illness from which he recovered after some months. He resumed his legal post but after eight years fell ill again. This time he remained in an asylum for nine years ( 1893-1902) towards the end of which he wrote his memoirs. In 1903, shortly after his discharge, they were published, and soon became a subject of psychiatric interest. A principal theme concerns a number of bodily experiences that were extremely painful and humiliating to him. These experiences he construed as 'miracles' performed by God by means of 'rays':
From the first beginnings of my contact with God up to the present day my body has continuously been the object of divine miracles. . . . Hardly a single limb or organ in my body escaped being temporarily damaged by miracles, nor a
-174-
single muscle being pulled by miracles, Even now the miracles which I experience hourly are still of a nature to frighten every other human being to death. 1
Freud's analysis of Schreber's delusions of persecution takes account of no material except that of the memoirs. Freud notes that Schreber's feelings towards God are intensely ambivalent, being on the one hand critical and rebellious and on the other reverential towards someone of whom he stands in awe. Freud calls attention also to the frankly homosexual attitude Schreber sometimes adopts towards God, including Schreber's belief that he had a duty to play the part of a woman for God's enjoyment. From material of this kind Freud postulates that delusions of persecution are attempts to contradict the proposition 'I (a man) love him (a man)', and to replace it by 'I do not love him -- I hate him', and, finally, by 'I hate him, because he persecutes me'.
140
An internal perception is suppressed, and, instead, its content, after undergoing a certain kind of distortion, enters consciousness in the form of an external perception. In delusions of persecution the distortion consists in a transformation of affect; what should have been felt internally as love is perceived externally as hate.
To this process Freud gives the name projection SE 12: 63-6).
In his re-examination of the case Niederland ( "1959"a and b) draws attention to the fact that Schreber's father held extraordinary views about the physical and moral education of children and published a number of books describing his methods. In them he asserts the vital importance of starting the prescribed re? gimes during infancy and states repeatedly that he has applied his methods to his own children. It is safe, therefore to conclude that Schreber the son had been subjected to his father's educational methods from his earliest years.
The physical methods, recommended for application daily throughout childhood and adolescence, include a number of exercises and harnesses whereby posture is to be controlled. An
____________________
1 In addition to Freud's paper and Strachey's editor's note in the Standard Edition, an English
translation of the memoirs is now available and also a paper by Baumeyer ( 1956) in which he summarizes and quotes from the original case records of Schreber's illnesses. Niederland's bibliography gives references to the above and to the published works of Schreber's father.
-175-
example of a harness, designed to prevent a child's head from falling forwards or sideways, consisted of a strap clamped at one end to the child's hair and at the other to his underwear so that it pulled his hair if he did not hold his head straight. Because the device was apt to produce a stiffening effect it was recommended that its use be restricted to one or two hours a day. An example of an exercise is to place two chairs facing each other with a gap between of a few feet.
A child is instructed to put his head on the seat of one chair and his feet on that of the other and to stiffen his back to make a bridge, in which position he must remain. The dire results that Schreber senior ascribed to bad posture included impeded circulation and, later, paralysis of arm and foot. Of one of his devices, an iron crossbar designed to ensure that a child sits straight, he comments that, besides its physical benefits, it provides an effective moral corrective.
Schreber senior held the sternest of views regarding moral discipline. Bad elements of the mind he regarded as 'weeds' to be 'exterminated', and he describes the threats and punishments by which, starting at five or six months, a parent should make certain that he becomes 'master of the child for ever'. The strong impression given that Schreber senior was a psychotic character is supported by a note made by a hospital psychiatrist and based, it is thought, on information from a member or close acquaintance of the family. It states that the patient's father 'suffered from obsessional ideas with murderous impulses'.
Niederland compares the son's descriptions of the fearful 'miracles' he had to suffer at God's hands with the father's prescriptions of how children should be treated for their physical and moral welfare. Point by point the resemblances are traced. The son complains of miracles of
141
heat and cold. The father prescribes that, in order to toughen an infant, he should be washed in cold water from the age of three months and also subjected to various local cold applications. The son complains that his eyes and eyelids are the target of uninterrupted miracles. The father prescribes repeated visual exercises and advises spraying the eyes with cold water should there be irritation and fatigue following over-stimulation. The son describes a miracle in which his whole chest wall is compressed. Father prescribes a harness consisting of an iron bar that presses against the collar bones should a child not sit straight and upright.
In view of these remarkable resemblances, Niederland's hypothesis, in keeping with the second of the four possible explanations listed above (pp. 173 -4), is that Schreber's delusory
-176-
beliefs regarding the way God was treating him were derived from memories of how his father actually treated him when he was a child. The delusory character of the beliefs is then regarded as due to (a) the patient's attributing the origin of his sufferings to the activities of God in the present instead of to his father in the past, and (b) his attributing the mechanism of his sufferings to 'rays' and miracles instead of to actual manipulations of himself by parent figures. As Niederland ( 1959a) himself remarks, the hypothesis is in keeping with ideas that Freud was entertaining towards the end of his life (but which still have been little exploited). In hallucinations, Freud ( 1937) suggests, 'something that has been experienced in infancy and then forgotten returns. . . . ' .
If this approach to understanding paranoid delusions is adopted, many problems remain still to be solved. How comes it that the patient has no recollection of how his parents treated him as a child? Why is it that, instead, childhood experiences are misplaced in time and the agent responsible for them is misidentified? Possible answers to these questions invoke hypotheses regarding the kinds of injunction, explicit or implicit, a parent may issue to a child; for example, an injunction on a child to construe whatever happens to him as beneficial, an injunction to see his parent as above criticism, an injunction neither to perceive nor to remember certain acts that he none the less witnesses or experiences. These hypotheses, with much evidence to show that they apply to the case of Schreber, are advanced in a recent paper by Schatzman ( 1971). Yet a further hypothesis, not discussed by Schatzman, is that children wish to see their parents in a favourable light and often distort their perceptions accordingly.
In Chapter 20 these matters are pursued further. Meanwhile, enough has been said to show that, when the actual experiences they have had during childhood are known and can be taken into account, the pathological fears of adult patients can often be seen in a radically new light. Paranoid symptoms that had been regarded as autogenous and imaginary are seen to be intelligible, albeit distorted, responses to historical events.
-177-
142
Chapter 12
Fear of Separation
Hypotheses regarding its development
It is now time to draw together ideas regarding fear of separation and how it develops. At the end of Chapter I it is pointed out that 'presence' and 'absence' are relative terms that can give rise to misunderstanding. By presence is meant 'ready accessibility', by absence 'inaccessibility'. The words 'separation' and 'loss' as used in this work imply always that the subject's attachment figure is inaccessible, either temporarily (separation) or permanently (loss). Thus in what follows we are concerned with the developmental processes that lead a young child to respond with fear when he finds, or believes, his attachment figure to be inaccessible. Among the many questions raised and not yet answered are the following:
1. Is inaccessibility of mother in itself a situation that arouses fear in human children without its being necessary for any learning to have taken place?
2. Or is such fear elicited in an individual only after he has come to associate her inaccessibility with a distressing or frightening experience?
3. If the latter, what is the nature of such distressing or frightening experience, and by what type of learning does it become linked with separation?
Whatever the answers to these questions may be, because being alone carries an increased risk of danger, especially for young individuals and others who are weak, the fear response to inaccessibility of mother can usefully be regarded as a basic adaptive response, namely a response that during the course of evolution has become an intrinsic part of man's behavioural repertoire because of its contribution to species survival.
If that is so, there is no a priori reason to assume that fear elicited by mother's inaccessibility can be explained only in terms of an individual's having experienced something dis-
-178-
tressing or frightening while separated from her, an assumption that has commonly been made. On the contrary, it is entirely possible that the response to mother's inaccessibility develops during ontogeny without learning of any sort having to take place. Let us call this hypothesis A.
Whether or not hypothesis A applies in the human case remains an open question. For, as was emphasized repeatedly in the first volume, there are many forms of behaviour that, like this one, can usefully be classed as instinctive but that develop functionally only when the environment provides opportunity for learning of some specific kind to occur. In other words, to hold the hypothesis that fear behaviour in a situation of maternal inaccessibility is instinctive in no way rules out the possibility that learning of some kind is necessary for its development. All that such a view requires is that, when an individual is reared in the species's environment of evolutionary adaptedness, opportunity for the necessary learning is always present.
143
Reflection suggests that there are at least three hypotheses that are consistent with that proviso and merit attention. Let us call them hypotheses B1, B2, and B3.
The first, B1, is Freud's hypothesis of 1926 which postulates that fear of mother's absence results from an infant's learning that, when she is absent, his physiological needs go unmet, and learning, further, that this results in the accumulation within him of dangerous 'amounts of stimulation' which, unless 'disposed of', bring about a 'traumatic situation'. Since, moreover, the infant finds that, left to himself, he is unable to dispose of such accumulations, the danger situation that he comes intrinsically to fear is 'a recognized, remembered, expected situation of helplessness'. 1
Reasons for not adopting Freud's hypothesis will already be apparent. One is that it is embedded in a paradigm very different from the one adopted here (see Chapter 5). Another is that it seems to postulate a degree of insight into cause and effect that not only is improbable in an infant of a year or so of age but that we now know to be unnecessary to account for the findings. For the fact that so many of the responses shown by a human infant when separated from his mother are to be seen also in infants of non-human primate species demonstrates that
____________________
1 Quotations are from Inhibitions, Symptoms and Anxiety ( SE 20: 137-8 and 166). Freud's
theory is described more fully in Appendix I. -179-
it is quite possible for such responses to be mediated at a primitive and presumably infra- symbolic level.
Objections of a similar sort apply to the theories advanced by Klein, which presuppose even more sophisticated cognitive functioning (see Appendix I).
A second hypothesis, B2, not very different from Freud's but simpler and implying no insight learning, is compatible with the theory of attachment behaviour proposed in the first volume. In Chapter 14 of that volume an account is given of the conditions that terminate crying during the early months of life:
. . . when a baby is not hungry, cold, or in pain, the most effective terminators of crying are, in ascending order, sound of voice, non-nutritive sucking, and rocking. These findings readily explain why babies are said to cry from loneliness and to have a desire to be picked up. Although to attribute such sentiments to babies in the early months of life is almost certainly not warranted, the statements none the less contain more than a grain of truth. When they are not rocked and not spoken to infants are apt to cry; when they are rocked and spoken to they cease crying and are content. And by far the most probable agent to rock and talk to a baby is his mother figure.
In view of this, it could be argued, an infant comes to learn that presence of mother is associated with comfort while absence of mother is associated with distress. Thus, through a fairly simple process of associative learning, an infant comes to associate mother's absence with distress, and so to fear her being inaccessible. This hypothesis is close to one advanced by Kessen & Mandler ( 1961).
144
A third hypothesis, B3, derives from the fact that an infant is much more intensely afraid of fear-arousing situations, such as strangeness, or sudden approach, or loud noise, when his mother is absent than when she is present. After a few such experiences, it could be postulated, mother's absence might of itself come to elicit fear, again through a process of associative learning. This hypothesis is similar to one suggested by Rycroft ( 1968a) and referred to in Chapter 6 above.
On present evidence it is not possible to decide between hypotheses A, B2, and B3; each is plausible.
Hypothesis A, that an anxiety response to mother's inaccessibility develops during ontogeny without learning of any
-180-
sort having to take place, is difficult to test. Furthermore, even if it were true, it would not make hypotheses B2 and B3 irrelevant, since learning of the kinds proposed by these two hypotheses could still occur and might be of much significance in accounting for degrees of separation anxiety above a minimum.
Whether hypothesis A is valid or not appears, at present, to be of no great clinical importance. This is because, were the forms of learning postulated by hypotheses B2 and B3 to occur at all, which they probably do, they would be taking place during the latter half of the first and during the second year of life and, except where a child had no mother figure, would be virtually unavoidable. As a situation that arouses fear, therefore, separation from an attachment figure would still be nearly universal, almost as much as it would were hypothesis A to apply.
Support for the view that associative learning of the kinds postulated by hypotheses B2 and B3 does take place comes from studies of individual differences in susceptibility to respond with fear, especially to separation. These show, as is discussed in detail in later chapters, that children who have been well mothered, and therefore, in all likelihood, have been protected from the experience both of intense distress and of intense fear, are those least susceptible to respond with fear to situations of all kinds, including separation; whereas children who have had intensely distressing and frightening experiences when away from mother are apt to show an increased susceptibility to fear, especially to fear of being separated again.
Should, as therefore seems likely, both these forms of associative learning occur during infancy and early childhood, their effects on personality development might possibly be rather different. For example, were a child, because of his particular experiences, to come to associate mother's absence with high degrees of discomfort and distress, he might perhaps grow up to respond to separation and loss, either actual or forecast, with psychosomatic troubles and general tension; whereas a child who, because of his particular experiences, came to associate mother's absence with being more or less intensely afraid might grow up prone to respond to any fear-arousing situation with more marked fear than would other individuals. Whether or not differences of these kinds occur in fact can be determined only by further research.
-181-
145
Need for two terminologies
Throughout recent chapters a sharp distinction has been drawn between situations that arouse fear and situations that are intrinsically dangerous. Whereas situations that arouse fear can be regarded as constituting either natural or cultural clues to an increased risk of danger, they are certainly not infallible indicators of actual danger. How we feel in a situation bears therefore only an indirect relationship to the degree of risk present in that situation. Because the world as reflected in feeling is distinct from, though correlated with, the world as it is, two terminologies are necessary. At the end of Chapter 6 three terms, 'anxious', 'alarmed', and 'afraid' are introduced and the way in which they are used here is described. All three belong to the world as reflected in feeling. By contrast, 'dangerous' belongs to the world as it is. At this point it is necessary to settle on some analogously distinct terms suitable to refer, on the one hand, to a state of feeling antithetical to feeling afraid and, on the other, to a situation antithetical to one of danger. Etymology suggests 'feeling secure' for the one and a 'situation of safety' for the other. The original meaning of the English adjective 'secure' is 'free from care, apprehension, anxiety or alarm' ( Oxford English Dictionary). Historically, therefore, 'secure' applies to the world as reflected in feeling and not to the world as it is. By contrast, the original meaning of 'safe' is 'free from hurt or damage'. As such it applies to the world as it is and not to the world as reflected in feeling. The distinction is neatly illustrated by a seventeenth-century saying, quoted in the OED, 'The way to be safe is never to bee [sic] secure', namely feel secure. By using the terms in their original senses, it is possible accurately and without ambiguity to make statements such as:
--although the situation was safe enough he became very frightened, or
--I could see the situation was dangerous but somehow the captain's behaviour made us all feel secure.
The distinction drawn here between feeling secure and being safe is not always made so that a number of terms current in the literature do not conform to the usage proposed. This applies both to Harlow's 'haven of safety', termed here 'secure base',
-182-
and to Sandler's 'feeling of safety' ( Sandler 1960), termed here 'feeling of security'.
Use of the word 'secure' in the sense proposed has, of course, for long been customary in clinical practice. For example, with reference to states of feeling, children and grown-ups are habitually described as being either secure or insecure. Moreover, because any person who is acting as an attachment figure for another is commonly referred to as providing that other with a sense of security, it is often convenient to describe an attachment figure also as a security figure or as providing a secure base. At the same time, it must be emphasized that a secure base, however much it may lead someone to feel secure, is no guarantee of safety, any more than a natural clue, however frightening we find it, is a certain indicator of danger. As a guide to what is safe and what is dangerous the kind of feeling a situation arouses in us is never more than rough and ready.
-183-
146
Part III
INDIVIDUAL DIFFERENCES IN SUSCEPTIBILITY TO FEAR: ANXIOUS ATTACHMENT
-185-
Chapter 13
Some Variables responsible for Individual Differences Constitutional variables
That individuals differ enormously in their susceptibility to respond to situations with fear is a commonplace. Why they should differ in such extreme ways remains a puzzle. In this chapter and those following an attempt is made to identify some of the many variables that are operating. The main focus, of course, is on the part played by a person's relationship to his attachment figure(s). This is held to be pervasive and still too little understood. Let us consider first some of the other variables. It must be assumed that genetic differences play some part in accounting for variance between individuals with regard to susceptibility to fear. Very little is yet known about their role in humans, but it is well documented in the case of other mammals, e. g. dogs ( Scott & Fuller 1965; Murphree, Dykman & Peters 1967). A difference in susceptibility in humans that is likely to be in part genetically determined is one between men and women.
Sex Differences
Feminist opinion notwithstanding, it is very commonly believed that there are some differences in susceptibility to fear as between men and women. This view is plausible and there is some evidence to support it. At the same time it is clear that in this regard there is much overlap between any population of women and a comparable population of men. Culture, moreover, can either magnify such potential differences as there may be, for example by sanctioning the expression of fear by members of one sex but not by those of the other, or else try to reduce them. Evidence from four sources supports the idea of a difference in susceptibility between the sexes:
In the experiments with nursery-school children, carried out by Jersild & Holmes ( 1935a) and described in Chapter 7,
-187-
a higher percentage of the girls were afraid than of the boys. The situations in which the difference was most marked were going into the dark passage and approaching the two animals, snake and dog. In these three situations the percentages of boys who showed fear were respectively 36, 40, and 46. The comparable percentages for girls were 48, 50, and 59.
In interviews of mothers of children aged six to twelve years Lapouse & Monk ( 1959) found that the proportion of girls reported as being afraid of strangers and animals, notably snakes, was higher than that of boys. In two other studies in which children of about the same age were interviewed, girls reported more situations as feared than did boys ( Jersild, Markey & Jersild 1933; Croake 1969).
147
In questionnaires given to students there is a consistent tendency for women to report more situations as feared than men (for references and comment see Marks 1969).
In epidemiological studies of psychiatric casualties women are reported to suffer from anxiety states about twice as frequently as men ( Leightonet al. 1963; Hare & Shaw 1965). Two-thirds of agoraphobic patients seen by psychiatrists are women ( Marks 1969).
A difference in the opposite direction -- that females tend to show less fear than do males -- seems not to have been reported.
Viewed in an evolutionary perspective these findings are not surprising. In most races of man, as in other species of groundliving primates, males are larger and stronger than females ( Cole 1963). While males bear the brunt of defence against predators, as well as attacking them when necessary, females protect young and, unless prevented from doing so, are more likely to retire from dangerous situations than to grapple with them. It would be strange were such long-standing differences between the sexes in respect of body structure and social role not to be reflected in complementary differences in behavioural bias.
Minimal Brain Damage
In Chapter 16 of the first volume an account is given of a longitudinal study of twenty-nine pairs of boys ( Ucko 1965), which shows that children who at birth are noted to be suffering from asphyxia are much more sensitive to environmental change than are matched controls. When the family went on holiday
-188-
or changed house, boys who had suffered from asphyxia were more likely to be upset than were the controls. The same was true when a member of the family -- father, mother, or sibling -- was absent for a time. These differences were apparent during each of the first three years of life (though not significantly so during the third). A comparable difference was seen when some of the children started nursery school.
Soon after his fifth birthday every child started infant school, making this the only event that was common to them all (though of course they went to many different schools). Here again the difference between the two groups was striking and significant. On a three-point scale (reduced from five points), the children distribute as shown below:
Enjoyed start accepted it
Mild
protest
within one week
Mild
school from the
or
at least
apprehension and disappearing
8 10 apprehension or 13 2
Asphyxiated at birth
Controls
8 17
148
marked
more than a week
Totals
Childhood Autism
disturbance
lasting
The behaviour of an autistic child shows a complete absence of attachment together with many indications of chronic fear. Tinbergen & Tinbergen ( 1972), adopting an ethological approach, suggest that the underlying condition may be one of chronic and pervasive fear, which cannot be allayed by contact with an attachment figure because the child also fears humans. If this is so, the syndrome could be conceived as resulting from a persistently lowered threshold to fear-arousing stimuli combined with delayed development of and/or inhibition of attachment. Causal factors might then include any of
-189-
the following: (a) genetic factors, (b) brain damage, (c) inappropriate mothering. A combination of two or more factors seems likely. Clancy & McBride ( 1969) describe a treatment programme based on this type of theory.
Blindness
Nagera & Colonna ( 1965) report that blind children are apt to be more than usually afraid of such common fear-arousing situations as animals, mechanical noises, thunder and wind, and to live in a state of permanent alertness. A principal reason for this is probably that, being blind, they are likely to be out of contact with their attachment figure far more often than are sighted children, and thus often to be effectively alone when something frightening occurs. Their tendencies on some occasions to remain rigidly immobile and, on others, to seek very close bodily contact with an adult are in keeping with this explanation.
Great difficulties arise for such children after a brief separation because a blind child cannot track his mother visually and keep close to her as a sighted child commonly does on such occasions. Fraiberg ( 1971) describes the very acute reaction of a blind boy of fourteen months after his mother had been absent for three days, during which he had been cared for by various friends and relations. During the first fortnight after mother's return he screamed for hours at the highest pitch, 'something between terror and rage', or else shouted and chanted perpetually. Only when his mother held him was there any respite; and then he would crawl relentlessly all over her. Because the screaming was so distressing to mother it was suggested she give him pots and pans to bang together instead. This the child did with great gusto and the screaming ceased.
Fraiberg describes also another blind child, a little older, who was cared for by familiar grandparents while mother had a new baby. When reunited with his mother he was markedly ambivalent at first but responded quickly when she, an affectionate mother, gave him plenty of cuddling. The main reason for the far more acute reaction in the younger child is likely to have been that his mother was a disturbed woman whose mothering was erratic both before
Asphyxiated at birth
Controls
29 29
149
and after her absence; another factor may have been that he was cared for by several different people while she was away.
-190-
Changes during Development in a Child's Susceptibility to Fear
While every infant comes into the world with biases to respond in some ways more than in others, how he develops turns on a process of interaction between himself and his environment. In regard to a susceptibility to respond fearfully, there are certain developmental trends sufficiently buffered to environmental variation to be seen in a huge majority of individuals. For example, as related in Chapter 7, all descriptive studies agree that, whereas during the first two years of life a child is broadening the range of situations he fears -- to include especially strangeness, animals, darkness, and separation -- from his fifth birthday onwards, and often before, he is likely to become steadily more discriminating in what he fears and more confident and competent in dealing with situations that would formerly have frightened him. Because change towards greater discrimination and confidence represents the norm, we start by considering the nature of the experiences and processes likely to be responsible for it. Subsequently we consider experiences and processes that have an opposite effect, for example, those that interfere with the usual tendency for susceptibility to diminish, or even enhance the susceptibility, and others that have the effect of increasing the range of situations feared.
Experiences and processes that reduce susceptibility to fear
The experiences that occur and the processes at work during the ordinary course of a person's life that tend to reduce his susceptibility to fear are of many kinds. A principal process, increasing confidence in the availability of his attachment figure(s), is the subject of the next chapter. Of the others the main ones can be described, in everyday language, as getting used to situations that are initially alarming, discovering that in many such situations other people are not afraid, and learning to tackle a situation actively and thereby discovering that nothing ill befalls. In the language of learning theory they are termed:
-- habituation
-- observational learning leading to vicarious extinction
-- observational learning combined with guided participation.
Other processes are likely to be at work as well, though it is not clear how large a part they play during the ordinary course
-191-
of development. For example, it may well be that there is some naturally occurring version of the procedure developed by behaviour therapists, and known variously as 'reciprocal inhibition', 'counter-conditioning', and 'desensitization', in which an association is gradually built up between a stimulus situation that is feared and something that the subject finds pleasant. 1
Yet another process, and one that it is easy to forget, is that as an individual grows up he becomes stronger and more skilful so that situations that might once have been, or at least seemed, dangerous to him cease to be so.
Knowledge of some of these processes has been greatly extended in recent years by the work of learning theorists and behaviour therapists. As Marks ( 1969) is at pains to stress, a large
150
majority of these studies have been conducted with healthy individuals who happen to be intensely afraid of some delimited object or situation, such as a snake or a dog, and not with psychiatric patients who commonly suffer not only more generalized anxiety but usually difficulties in personal relationships and a tendency to depression as well. It is for this reason that many clinicians suspect that the findings of the learning theorists may prove to be of only limited value in psychiatric practice. Yet it is for this same reason that their findings are in all likelihood of much relevance in understanding how it happens that the tendency to respond fearfully recedes during the ordinary course of healthy development.
Let us consider further the three processes already listed.
Habituation
This is a process of learning not to respond to a situation when it is followed by nothing of consequence. It presumably plays a major part in restricting an infant's initial tendency to respond with fear to all and any strong or sudden stimulation. Later, habituation, perhaps in more sophisticated forms, also restricts the range of situations that are responded to with fear because they are strange; for much of what is strange today will not only have become familiar tomorrow but also have been found to lead to no untoward consequence. Thus habituation greatly limits the range of situations responded to with fear. It should none the less be noted that habituation in no way affects
____________________
1 A full description of desensitization and related techniques is given in Marks ( 1969).
-192-
the basic and persistent tendency to respond with fear, as well as with curiosity, to anything perceived as strange.
Observational Learning leading to Vicarious Extinction
It has already been remarked that observational learning can work in either of two directions: either the observer learns to fear situations that formerly he did not fear or else he learns not to fear situations that formerly he did fear. The most important component in learning not to fear situations formerly feared, Bandura ( 1968) finds, is that the observer should see that the feared situation can be approached and dealt with without there being any bad consequences. The identity of the person observed (model) and the degree to which the observer can identify with him are found to be of much less significance. Even watching a sequence on film can have a reassuring effect, provided always that the consequences of the model's actions are clearly depicted.
The process of learning that something is harmless from direct observation of the experience of others is very different, it should be noted, from merely being informed by another person that a situation is harmless. All those who have made a systematic study of the problem report that simple explanation and reassurance have only very limited effect, a finding that will come as no surprise to clinicians.
Fortunately, in the ordinary course of events, a child growing up in a family has endless opportunities to learn from observation that many of the situations that make him afraid are in
151
fact harmless. Parents, older brothers and sisters, neighbours and schoolfellows are continually and without knowing it providing a child with this indispensable information.
Observational Learning combined with Guided Participation
This method requires much more from the model than giving the subject opportunity for simple observational learning. It is evident none the less that every sensible parent is constantly providing it. The method consists in the model's first demonstrating in action that the feared situation holds no danger and then encouraging the other person -- child or adult -- to tackle the situation himself. Once again it appears that the crucial part of the process is that the learner should discover, this time for himself, that approaching and tackling the situation can be done without untoward consequences.
