By contrast, to respond to the natural and
cultural
clues is quick and simple.
Bowlby - Separation
Why waste time on these primitive mechanisms?
The reason is, of course, that much of the sophisticated superstructure of cognitive and feeling processes characteristic of Western man in the realm of fear is intelligible only in terms of the primitive genetically biased groundwork that evolved in a different environment and that we share with other primate species.
A failure to understand this primitive groundwork, it is argued, has led to many and serious misunderstandings.
Not only is the behaviour of every human adult influenced by these primitive processes but so also are his most sophisticated cognitive structures and his most sensitive ways of feeling.
-140-
Whether suddenly alarmed or chronically anxious, whether temporarily comforted or steadily confident, the way a man or woman thinks and feels is determined in significant degree by these strong genetic biases to respond unthinkingly to the natural clues.
In the chapters that follow attention is given, first, to showing that the strong tendency to respond to the natural clues accounts for most of the more elaborate situations that humans come to fear, and, subsequently, to the way in which increasingly refined processes of appraisal lead to a broad spectrum of human feeling states. Before proceeding, however, let us consider further the genetically biased groundwork. We begin with the special place of physical pain as a natural clue.
The Limitations of Pain as a Natural Clue
In the past there have been theorists who have postulated that almost the only type of stimulus to which there is any genetic bias to respond with fear is physical pain, and that all other stimuli derive their fear-arousing properties from becoming associated with pain. Not only is the theory false, but a moment's thought shows it to be hardly plausible.
115
As a natural clue to potential danger, the experience of physical pain is in a special category. The clues to which attention has so far been directed are distal clues perceived by the distance receptors, eye, ear, and nose. By giving warnings while potential danger is still more or less remote, these clues enable an animal or man to take precautions in good time. By contrast, as noted in the previous chapter, to await events until pain is experienced may well be to wait too long. Whereas the distance receptors can be likened to advance look-outs, physical pain has the status of last ditch.
The special property of pain is that, being so late in acting, it leads to immediate and urgent action. The phase of alert wariness, so characteristic of many animals after a distal clue is first sensed, is absent. Instead, there is immediate and unthinking retreat, or, alternatively, attack.
Another special property of pain is, of course, its power to promote learning. Countless experiments demonstrate how rapidly and firmly an animal learns to recognize a situation in which it has experienced pain and to respond thenceforth by avoiding it. After such learning, an animal no longer relies on the hazardous proximal clue of pain but comes instead to use some distal clue that gives time and space in which it can take
-141-
precautions. The advance look-outs are alerted to identify and beware of a new clue.
Even though physical pain may be more highly correlated with potential danger than are some of the other natural clues, it is not infallible. For example, medical attention may be painful but is usually not dangerous; whereas a truly dangerous condition, such as internal haemorrhage, may be accompanied by no pain. That is but one example of a serious danger that is either without natural clues or heralded by faint ones only.
Dangers that have no Natural Clues
Earlier it was noted that the natural clues to which we react with fear are, singly and especially together, indicators of a high proportion of all the dangerous situations into which we might stray. Nevertheless, there are some dangerous situations that present no clue to which we have a natural bias to respond by escape. Some, indeed, even emit no signal that our sense organs can detect.
Among naturally occurring hazards, infectious illnesses are cases in point. Where infection is airborne there is usually no naturally occurring clue that we are able to sense and from which we are genetically biased to withdraw. (In contrast, by producing bad smells or tastes, food- and water-borne infections are often much less silent. ) In modern times, moreover, man has added a number of other dangers that also emit no clues to which human nature is sensitive. Examples are carbon monoxide gas and X-rays. Since in such cases evolution has as yet had neither time nor opportunity to provide us with natural means for their detection, we have to rely instead on man-made indicators.
Thus, although by exploiting the natural clues to danger and safety our genetic endowment provides us with a remarkably sensitive and efficient means of protection, it is far from foolproof. On countless occasions we are led unnecessarily to avoid wholly harmless situations; on a few others we are permitted to blunder into truly dangerous ones.
116
Potential danger of being alone
The natural clue to increased risk of danger with which this volume is especially concerned is being alone. Statistically, being alone is less safe than being with a companion. That that should be so during childhood, during sickness, and in old age
-142-
may not be too difficult to grasp. That it is also the case for the ordinary healthy grown-up man and woman may at first sight seem unexpected. Yet there is good reason to believe that it is so, especially in certain situations; even though in Western countries the situations may be few and the absolute risk not high. The thesis of this section is, therefore, that, as it was throughout man's earlier history, in many circumstances still today it is as appropriate to avoid being alone as it is to avoid any of the other natural clues to potential danger. That we should be so constructed that we find comfort in companionship and seek it, and that we experience greater or less degrees of anxiety when alone, is, therefore, in no way surprising.
In the previous volume (Chapter 4) it is argued that, if we are to understand human behavioural equipment, it is necessary to view it in the light of what we know of man's environment of evolutionary adaptedness. Later, following this line of thought (Chapter 12), the theory is advanced that in man's environment of evolutionary adaptedness the function of attachment behaviour, which of course promotes proximity to special companions, is protection from predators, and that this is as true for humans as it is for other species of mammal and bird. For all ground-living primates, safety lies in being with the band. To become separated from it is to provide a more or less easy meal for a lurking leopard 1 or a pack of hunting dogs. For weaker members, especially females and young, the old and the sick, isolation often spells speedy death.
By practical people this theory is sometimes treated as an academic curiosity. Yes, it may be said, there may well have been a time in man's history when separation entailed danger from predators. But that was long ago. For such responses to persist into modern times is an irrelevant nuisance. It is time to rid ourselves of such archaic superstition.
This line of reasoning has several defects. In the first place, even if we wished it, genetic biases built in over millions of years cannot be eradicated overnight. In the second, reflection suggests that to try to eradicate them might be most unwise. For in many parts of the world today the absolute risk attendant on
____________________
1 Since publication of the first volume further evidence has come to light of the dangers of
leopards to early man. According to Brain ( 1970) the fossilized bones of Paranthropus robustus found in a cave in the Transvaal are fragmented in ways typical of leopard prey. One of the betterpreserved skulls (of a juvenile) bears two holes the right size and distance apart to fit the canine teeth of a leopard.
-143-
being alone is still fairly high; and even in Western societies the risk may be higher than we like to imagine.
117
In Western countries today, it is true, injury and death are no longer due to predators. But there are other dangers. In place of predators, power-driven motor cars and household equipment provide novel hazards and take their toll. Young children newly mobile and the elderly are among the principal victims. Yet, though ordinary experience strongly suggests that those most at risk are children and old people left on their own, researchers concerned with accident prevention seem to have given the matter little attention. Statistics of traffic accidents for one of the London boroughs and also for Sweden are, however, revealing.
Traffic Accidents to Children
During 1968 in the London borough of Southwark, 1 injuries to pedestrians numbered 901, of which twenty-seven were fatal. Of the total injuries, 411, or nearly half (46 per cent), were to children under the age of fifteen. This represents an incidence of injury to children about three times that to adults.
The most vulnerable age-group were children between their fourth and eighth birthdays. At these ages the risk of injury was about five times that for adults. The incidence for those a little younger and a little older (the three- and the eight-year-olds) was only slightly lower. The age distribution was as shown in the table below.
Age in No. of
years
0- 2? 11 3- 5? 11 6- 8? 11 9-11? 11 12-14? 11 Total
injuries
14
125
124
81
67
411
Of all the children injured, almost two-thirds (62 per cent) were entirely alone. Even in the case of the younger children
____________________
1 I am indebted to Mr V. E. Golds, Road Safety Officer for the borough, for these figures.
-144-
more than half were alone. Many of the others were with other children, often no older than themselves. Only one in eight of the injured children was with an adult.
A similar picture emerges from Sweden ( Sandels 1971). The incidence of injury to pedestrians is especially high in children between their third and tenth birthdays. A special study of 177 accidents occurring to children under the age of eleven years at pedestrian
118
crossings shows that 44 per cent of the children were alone and another 34 per cent with peers; only one in five was with an adult.
From these figures we can conclude that the great excess of traffic accidents to children when compared with adults is due to their being out in the street either alone or with peers. 1 To anyone who has had the care of small children in an urban district this conclusion will hardly cause surprise.
Risk to Adults
It is perhaps easy to understand that for a young child or an old person to be alone is a risk. But, it may be protested, that can hardly be true also for a healthy adult. Reflection, however, strongly suggests that it is.
It seems very probable that, were comparative figures available, it would be found that even for healthy men and women in Western countries there are many situations in which risk of injury or death is greater when a person is alone than when in company. Walking in city streets at night is a case in point. It is not for nothing that in certain areas policemen patrol in pairs. Those who take part in active sports, moreover, are aware that to be alone carries added risk. Whether climbing mountains, swimming, exploring caves, or sailing the seas, to be alone is hazardous, sometimes because in detecting danger two heads are better than one, sometimes because an injury that would present no problem to a pair can prove fatal to a singleton.
____________________
1 Studies of the family background of children who are injured in traffic accidents ( Backett
& Johnston 1959; Burton 1968) cast light on why these children are not being looked after by a parent. When compared with children in a control group, more of the injured children are found to be unwanted and unloved and/or to have a mother who, currently, is anxiously preoccupied with other matters, e. g. illness in herself or in others in the household, younger siblings, elderly relatives, or her own pregnancy. Similar findings for children who sustain burns are reported by Martin ( 1970).
-145-
Yet another hazard to a man alone comes when he is overtaken by fatigue. Once asleep he cannot protect himself should danger threaten. When, by contrast, he has companions, each can take his turn on watch. The practice of alternating watches on board a ship at sea is, indeed, the organized and human version of a sleeping pattern common in birds that roost together in flocks and in primates that sleep together in bands. Because every animal is awake for some part of the night, at any one moment, while the majority of animals are asleep, a few are likely to be awake ready to give the alarm ( Washburn 1966).
It is true that in recent years great feats of single-handed navigation have been performed. That interest in their success should be so high is an earnest of the public's recognition, not only of the difficulties to be solved, but of the risks to be surmounted. Safety lies in numbers, especially in the companionship of familiars.
119
Potential safety of familiar companions and environment
Throughout these chapters it is emphasized that what is feared includes not only the presence, actual or imminent, of certain sorts of situation but the absence, actual or imminent, of certain other sorts of situation. Throughout life we tend to be drawn towards certain parts of the animate and inanimate environment, mainly people and places we are familiar with, and to be repelled by certain other parts of the environment, especially those that exhibit one or more of the natural clues to potential danger. Since two of the natural clues that tend to be avoided are strangeness and being alone, there is a marked tendency for humans, like animals of other species, to remain in a particular and familiar locale and in the company of particular and familiar people.
It has long been obvious that animals of any one species tend to restrict their movements so that they remain within those parts of the earth's surface to which they are physiologically adapted. Such parts can be defined in terms of various physical measures, such as earth, air or water, temperature gradients, rainfall, and also in terms of biological measures, such as presence or absence of certain foodstuffs. Only by regulating their movements in these ways are members of a species able to maintain the physiological measures on which life depends within certain critical limits. The types of behavioural system, activation and termination of which result in an animal's re-
-146-
maining within its ecological niche, are of the kind traditionally termed instinctive.
Yet, great though ecologically determined limitations may be, they are nothing in comparison with the limits constantly found in nature. It is still too little realized, perhaps, that the individuals of a species, so far from roaming at random throughout the whole area of the earth's surface ecologically suitable to them, usually spend the whole of their lives within an extremely restricted segment of it, known as the home range. 1 For example, a vole lives within its few hundred square yards of thicket, a troup of baboons within its dozen square miles of savanna, a band of human hunters and gatherers within its few hundred square miles of forest or plain. Even flocks of migrating birds, which may travel thousands of miles between nesting and wintering grounds, use only special parts of each: many birds nest each year at or very near the place they were born.
In a similar way birds and mammals do not mix indiscriminately with others of their kind. Individual recognition is the rule. With certain individuals close bonds may be maintained for long stretches of the life-cycle. With a number of others there may be a less close but none the less sustained relationship. Yet other individuals may either be of little interest or else be carefully avoided. Thus each individual has its own relatively small and very distinctive personal environment to which it is attached.
While the survival value of an animal's predisposition to remain within an ecologically suitable environment is clearly not in doubt, the survival value of its strong tendency to remain within its own special and familiar environment may at first sight seem debatable. Yet
120
examination of the issue shows that to do so in all likelihood confers distinct advantage, especially when conditions turn unfavourable. By remaining within a familiar environment an animal, or a human, knows at once where food and water are to be found, not only at different seasons of the ordinary year but also during those exceptionally bad years that occur from time to time; he knows, too, where shelter from the weather can be got, where there are trees or cliffs or caves
____________________
1 The concept of the home range embraces that of territoriality but is much broader. Whereas
very many species of bird and mammal show marked preferences for a particular home range (see Jewell & Loizos 1966), far fewer maintain and defend an exclusive territory. For a discussion of the probable functions of territory holding, which may differ between species, see Crook ( 1968).
-147-
that provide safety, what are the common dangers and from what quarter they are likely to come. By remaining in company with familiar companions, he can benefit from established and therefore relatively successful customs, for example food preferences, and also, when threatened by a predator, from coordinated social action. Thus, by remaining within his personal and familiar environment an individual stays within an arena that is comparatively safe and is kept well clear of many hazards that might otherwise endanger him.
Maintenance of an individual within his familiar environment is, it is postulated, the result of the activation and termination of behavioural systems that are sensitive to such stimulus situations as strangeness and familiarity, being alone and being with companions. On the one hand, behavioural systems mediating fear behaviour tend to remove the individual from situations that are potentially dangerous. On the other, those mediating attachment behaviour tend to lead him towards, or retain him within, situations that are potentially safe.
This brings the discussion back to attachment behaviour. The behavioural systems that maintain a younger or weaker individual in more or less close proximity to another discriminated and stronger individual can now be seen as a part of a larger set of systems that have the effect of so regulating the whole of a creature's movements that he remains as a rule within his familiar environment. Attachment to a parent figure is, in most species, ontogenetically the first form in which this type of behaviour develops.
In the next chapter and in Chapters 18 and 19 it is argued that many of the difficulties that have dogged psychiatric and psychoanalytic theories of anxiety have arisen because insufficient recognition has been given to the enormous roles that an individual's personal and familiar environment, including his familiar companions, plays in determining his emotional state. Only when it is realized that each man's environment is unique to himself can how he feels be understood.
Maintaining a stable relationship with the familiar environment: a form of homeostasis
Those trained in physiology may find it illuminating to view the behaviour under consideration as homeostatic. Whereas the systems studied by physiologists maintain certain
121
physicochemical measures, internal to the organism, within certain limits, the systems mediating attachment behaviour and fear
-148-
behaviour maintain the whole individual within a defined part of the environment. In the one case the states held steady are interior to the organism, in the other the states held steady concern the relationship of the organism to the environment.
A principal advantage of casting the theory of fear and anxiety presented here in terms of homeostasis is that by so doing it becomes possible to relate it to two other bodies of theory, both of which commonly invoke homeostatic principles. On the one hand, the theory can be linked to theories of stress and stress diseases, most of which invoke concepts of physiological homeostasis. On the other, it can be linked to a theory of defensive processes, which also, traditionally, have been conceived as contributing to the maintenance of some form of homeostasis. In place, however, of Freud's postulate that defensive processes help to reduce the level of stimulation in the mental apparatus to a steady low level, the theory advanced here conceives them as contributing to the maintenance of what can conveniently be termed a steady 'representational' state. 1
In the view presented here, maintenance of a steady relationship between an individual and his familiar environment occurs hardly less automatically and unthinkingly than does maintenance of his physiological steady states. In the case of each form of homeostasis, an individual is conceived as being born with a strong genetic bias to develop biological systems that, by being sensitive to certain types of stimuli, come into action whenever some particular measure deviates from certain set limits and cease action as soon as it is restored within those limits. Thus the theory proposed places the maintenance of a steady relationship between organism and familiar environment on a level of biological importance only one step lower than the maintenance of the much better understood physiological steady states.
Furthermore, the systems maintaining each of the two forms of homeostasis are seen as complementary. For it is clear that, so long as the systems that maintain an individual within his familiar environment are being successful, the loads placed on the systems that maintain physiological states steady are being eased. This is because, so long as an individual remains within his familiar physical environment and with familiar companions,
____________________
1 A theory of defensive processes incorporating these ideas is to be sketched in the later part
of the third volume, in which also a comparison will be made with certain other concepts of homeostasis that from time to time have been proposed by psychoanalysts.
-149-
he is more likely than he would be otherwise to find food and drink, and to achieve reliable and continuing protection from natural hazards -- from predators, from eating poisonous foodstuffs, from falling and drowning, from exposure and cold. Conversely, so long as the systems maintaining physiological homeostasis are successful, the healthier will the individual be and the easier will it be for him to maintain himself effectively within his familiar environment. Looked at in this light the regulatory systems that maintain a steady relationship between an individual and his familiar environment can be regarded as an 'outer
122
ring' of life-maintaining systems complementary to the 'inner ring' of systems that maintain physiological homeostasis.
It must, of course, be stressed that, whatever category of homeostasis is being considered, states are never maintained more than relatively stable nor, except rarely, do set-points and limits persist unchanged throughout the life-cycle. So long as the unit of study is the individual, indeed, processes of growth are the antithesis of processes of homeostasis. The principle of homeostasis is, therefore, only one among several. The reason for emphasizing it here is that homeostasis is held to be a key concept for understanding not only fear and anxiety but also grief and mourning.
While special attention is directed to the tendency for an individual to maintain a steady relationship between himself and his familiar environment, the important roles of exploratory and investigative behaviour, which tend to be antithetical to it, are not overlooked (see Volume I, Chapter 13). Nor are the developmental changes that occur in regular procession during the life-cycle.
-150-
Chapter 10
Natural Clues, Cultural Clues, and the Assessment of Danger
Innately, children seem to have little true realistic anxiety. . . . They will run along the brink of the water, climb on to the window-sill, play with sharp objects and with fire -- in short, do everything that is bound to damage them and to worry those in charge of them. When in the end realistic anxiety is awakened in them, that is wholly the result of education; for they cannot be allowed to make the instructive experiences themselves.
SIGMUND FREUD ( 1917b)
Clues of three kinds
Initially during infancy the only stimulus situations to which a child responds with fear are the natural clues. During the second and third years other situations are added, notably presence of animals, and darkness (with its associated happenings), both of which situations, it is argued, are readily learnt derivatives of the natural clues. From his second year, in addition, a child is much influenced by observing the behaviour of significant adults and imitating it. Among the great array of behaviour a child learns in this way, all of it culturally determined, is to respond with fear to a range of stimulus situations he has hitherto treated as neutral, or even interesting. These new stimulus situations are conveniently referred to as 'cultural clues'. In many instances, it is evident, imitative fear behaviour of this sort is entirely without insight into the nature of the danger being avoided. Because of this, fear aroused by a cultural clue has much in common with fear aroused by a natural clue. In neither case can the fear be regarded as 'realistic' in Freud's sense of the term.
Only very slowly and as his cognitive capacities develop does a child begin to distinguish natural or cultural clues from real danger and to learn methods of his own for calculating risk. During the same phase of growth all his behaviour is becoming organized increasingly in terms of goal-corrected plans, and
123
-151-
with it his fear behaviour. As a result of these linked developments his fear behaviour is becoming, as it is said, more 'rational' and 'realistic'. Thenceforward, during later childhood and adolescence, and into adult life, his capacity to assess real danger and to respond appropriately is likely steadily to improve.
Nevertheless, important though these new developments in the organization of fear behaviour may be, the bias to respond with fear to both the cultural and the natural clues persists. Indeed, not only throughout childhood but throughout adolescence and adult life as well the natural clues and their derivatives remain among the most effective of all the stimulus situations that arouse fear. Even the most courageous is not immune to fear on seeing some extraordinary apparition or a sudden rapid approach, or on hearing some piercing scream or finding himself alone in darkness in a strange place.
In intellectual circles both the persisting bias to respond to the natural clues and the value of that bias are all too often overlooked. As a result much human fear comes to be seen in a false perspective. For example, Arnold ( 1960), rightly impressed by the role of appraisal in the regulation of behaviour, goes so far as to assert that 'genuine fear develops only when the child is old enough to estimate the possibility of harm'. Elsewhere, in discussions of fear in human beings, there is commonly to be found the assumption, more or less explicit, that whereas fear of real danger is a healthy and often desirable response fear of anything else is childish or neurotic. Throughout psychiatry that assumption has for long been powerful and persuasive. It is found not only in the psychoanalytic tradition from Freud himself onwards (see above, Chapter 5) but in other traditions of psychiatry also (e. g. Lewis 1967). It is a main reason why the fear of separation from a loved figure is still so often, and so erroneously, held to be both neurotic and childish.
A principal thesis of this work is that the assumption that mature adults are afraid only of real danger, plausible though it may seem, is profoundly mistaken. Naturally enough an adult man, or woman, does what he, or she, can to calculate the prospects of real danger and to take the necessary precautions. Yet to make those calculations is often far from easy, and on some occasions would take dangerously long to do.
By contrast, to respond to the natural and cultural clues is quick and simple. To respond to the natural clues, moreover, especially when two or more people are present together, provides, as described
-152-
in the previous chapter, an efficient, if crude, system for minimizing danger and maximizing safety. No wonder, therefore, that alongside the more sophisticated measures for calculating danger, adult man continues to respond, at least tentatively, to each one of the natural clues and, when faced with compound situations, to respond especially strongly. Thus, in adult man fear behaviour comes to be elicited by clues that derive from at least three sources:
-- natural clues and their derivatives
-- cultural clues learnt by observation
-- clues that are learnt and used in more or less sophisticated ways in order to assess danger and avoid it.
Behaviour based on clues of the first type develops very early and is apt to be referred to as 'childish' and 'irrational'. Behaviour based on clues of the third type develops much later and
124
is commonly referred to as 'mature' and 'realistic'. Behaviour based on clues of the second type is intermediate: whether it is referred to as childish or mature, rational or irrational, turns on whether the onlooker shares or does not share the cultural norm reflected in the behaviour. For example, fear of ghosts is judged realistic by an observer from one culture and as childish by an observer from another.
A true evaluation of behaviour based on these three distinct types of clue is held to yield a picture very different from the popular one. Behaviour based on clues of the first and second types no less than behaviour based on those of the third is plainly consistent with normal development and mental health. In a healthily functioning individual, indeed, responses to clues of all three types are present; they can occur simultaneously or sequentially, and be either compatible with one another or in conflict.
In this chapter we consider the role of behaviour elicited by each of these three types of clue. Because so much attention has already been given to the natural clues we begin by considering the more sophisticated methods of assessing and avoiding danger.
Real danger: difficulties of assessment
Psychiatrists often speak as though it were easy to assess real danger. This is not so.
Both in ordinary life and in clinical practice there are two distinct types of problem. One is the difficulty that each of us has in assessing what is and what is not a real danger to his
-153-
own interests. Another is the difficulty that each of us has in assessing what is and what is not a real danger to another person.
Difficulties are met as soon as we try to define what we mean by 'real danger', whether to self or to another. There are a number of problems. One concerns how widely each of us draws the boundary of where his interests lie. A second concerns our understanding of what can, or cannot, cause injury. A third concerns the very varying ability of individuals to protect themselves and their interests: whereas a strong man might well be able to protect himself in a certain dangerous situation, a weaker one, or a woman or a child, might not.
We start with the problem of where each draws the boundary of his interests. Plainly, any situation that might lead to our own injury or death is classifiable as dangerous. The same would be agreed for anything that threatens injury or death to members of our family and to close friends. Beyond that, definition becomes more difficult. How widely do we extend the circle of friends and acquaintances whose safety we are concerned about? To what extent do we identify ourselves with the safety and wellbeing of the institution in which we work or the recreational club to which we belong? How do we rate threat to personal possessions, to house, and to favourite haunts?
Experience shows that a human being is constantly made afraid and anxious by threats of damage to a circle of persons, possessions, and places some way beyond himself and his body. For that reason it is necessary to include within the concept of real danger threat of injury or damage not only to the person himself but to the whole of his personal environment, as it is defined in the previous chapter.
125
All too often the need to include within the boundary the whole of an individual's personal environment is not recognized or, even if the principle is recognized, the nature and extent of the personal environment of a particular individual are not properly known. As a result, what is truly a danger for that person may go unnoticed by an onlooker.
Furthermore, not only is the nature of threat strictly relative to the person concerned but, as already remarked, means of protection are so too. Strong and competent people are able to protect themselves in situations in which weaker and less competent ones cannot.
Even when a definition of real danger is agreed, however, there remain great difficulties for each of us in assessing it.
-154-
For example, for an individual to calculate accurately when and in what degree he and his interests are endangered requires him to have a comprehensive knowledge of the world about him and to be able reliably to predict results. How many of us are qualified in these respects? It is easy to talk of real danger, but very difficult to estimate it.
It is indeed easy to forget that what is held to be publicly and permanently real is never more than some schematic representation of the world that happens to be favoured by a particular social group at a particular time in history. To some people during some periods to be afraid of ghosts is realistic. To other people during other periods to be afraid of germs is realistic. In matters of reality we all stand in danger of being arrogantly parochial.
That, however, is not to assert that everything is subjective, that there is no reality. The difficulty in using reality as a criterion lies, not in there being no reality, but in our imperfect capacity to comprehend it. That a child has an imperfect capacity to comprehend what is or may be truly dangerous is usually taken for granted. That the capacity of an adult is greater often by only a small margin tends to be forgotten.
To assess a risk of danger accurately requires us to take into account simultaneously a number of factors. Consider, for example, how we calculate the risk of being attacked by a particular dog. The ordinary dog, it will be agreed, is a harmless and amiable creature. Yet some dogs are dangerous to some people sometimes. What then are the criteria to be applied? On reflection we realize that they are numerous and complex. An accurate forecast rests partly on the sort of dog it is, partly on the situation in which we encounter it, partly on its behaviour and partly on how we estimate our own strength. Thus, we need to take into account the dog's age and sex, its breed, and perhaps also its probable training. Simultaneously, we should take into account whether the dog is on its home ground or elsewhere, whether with or without its master, and whether, in the case of a bitch, it has puppies. At the same time we should consider whether the dog is familiar with us, how it greets us, and how effective we judge ourselves to be in countering threat with threat and protecting ourselves should it attack. It is in fact a complex appraisal requiring considerable knowledge of dogs and accurate perception of the current situation. No wonder many adults as well as children despair of making it and behave as though all dogs are dangerous until
-155-
126
proved safe. Others, simplifying the situation in an opposite direction, may make the opposite assumption.
Consider again the difficulty of estimating accurately the danger of food-poisoning. To do so requires intimate knowledge of the food's origin, who has handled it, whether or not it has been cooked and the capabilities of various organisms to survive heating to various temperatures and for varying lengths of time. No wonder the ordinary housewife bases her behaviour on a limited number of culturally derived clues and practices.
In his capacity to assess and forecast real danger a child is of course even worse placed than an adult. Not only is he likely to be ill informed but, as Piaget has repeatedly shown (see Flavell 1963), his capacity to take into account more than a single factor at a time grows only slowly. It is fortunate that a child responds so readily to the natural and the cultural clues. Were he not to do so he would soon be dead.
'Imaginary' dangers
Assessment of danger always takes the form of a forecast. Sometimes the dangerous situation foreseen is judged to be imminent, at other times to be remote. In either case the likelihood of the situation's eventuating is of every degree. Dangerous situations that almost every adult in a society forecasts as 'probable present no problem. It is those situations that almost every adult forecasts as highly improbable or even impossible that are the challenge. Scoffingly, fear arising from such forecasts is dubbed 'exaggerated' or 'imaginary'; in more sober vein it is termed 'inappropriate'. For long, fear exhibited about such possibilities has constituted one of the principal riddles of psychopathology.
Yet, once the difficulty of making accurate forecasts of danger is grasped and once it is realized that if living beings are to survive there can be no great margin for error, the so-called imaginary fears come to be seen in a different and more sympathetic light. That children, still with a very imperfect model of the world, should at times gravely underestimate a danger may at times alarm us but is no surprise. That they should as frequently make an error of the opposite kind, foreseeing danger when we foresee none, is, when viewed in this perspective, no surprise either. Thus, when the bathwater goes down the plughole, how is a toddler to know he will not go down too? When, later, he hears tales of robbers and red indians intercepting coaches or robbing mail-trains, how is he
-156-
to know that he and his family may not be the next victims? The very great difficulty a child has in appraising at all accurately the degree of danger in which at any moment he may stand accounts, it is argued, for a much larger proportion of the so-called imaginary fears of childhood than is often supposed.
Sometimes 'imaginary' fear arises because of a simple misunderstanding, as when a small boy of six and a half, acting as a photographer's model, ran hurriedly off stage each time the photographer was about to press the button. Not until the next day was it revealed that, as soon as he heard the word 'shoot', he had run for his life. A similar type of misunderstanding led a boy of twelve, referred for stealing, to insist on having a sixpence in his pocket when he came to the clinic. The mystery was solved after some weeks when it transpired that he
127
believed the clinic to be a penal establishment, and that he had plans that if incarcerated he would escape, and he would then need the money for his bus fare home.
At other times 'imaginary' fear is a consequence of generalizing from too small a sample. If granny can die today, perhaps mother or father may die tomorrow. If a mother's first infant has died, is it surprising that she fears her second may die also?
The examples so far given are of erroneous or disproportionate forecasts of danger that arise from inaccurate or inadequate data. Until the source of an individual's erroneous forecast is known, his tendency to fear a particular situation will appear to another person to be absurd; it will also persist. Once the source is known about, however, the tendency is quickly seen by the other person to be far from unreasonable, even if misguided; and there is then a chance of its being corrected or modified.
In other cases fear of a situation that it may seem ridiculous to an outsider to fear can be explained in other ways. One source of such fear that has been greatly underestimated in the clinical literature is a forecast of danger that is in fact well based but remains inexplicable to an outsider because it is derived from information of the greatest secrecy. An example is seen in a child or adolescent one of whose parents is given to uttering dire threats -- of suicide, leaving home, even murder -- during emotional outbursts that, though real enough at the time, may be infrequent and in general out of character. While child or adolescent, not unnaturally, takes the threat seriously, the notion that such threats could ever be made may be discounted
-157-
or even denied by the parents. The key role that such family situations can play in accounting for the greatly intensified degree of separation anxiety suffered by some patients is considered in later chapters.
Another source of apparently unreasonable fear is a forecast of danger that derives from an individual's knowledge, conscious or unconscious, of certain desires of his own; for example, hostile wishes directed against someone he loves. Here again, to be afraid ceases to be unreasonable once the facts are known.
Yet other sources of fear that is or appears to be ill based lie in processes of projection and rationalization, to which brief attention is given in the next chapter.
In Chapters 18 and 19 further attention is given to some of the so-called irrational fears of anxious children and adults. These few paragraphs are intended to show only that the theoretical approach adopted can encompass without difficulty clinical problems of the greatest concern to every practising clinician and that a biological perspective in no way negates Freud's profoundly important discovery that fear can arise, not only from forecasts of how the external world and the people in it may behave, but also from forecasts of how we ourselves may possibly act.
Perhaps the most fundamental lesson to be learnt by anyone who wishes to understand the situations that other people fear is that forecasts of future dangers are as often as not strictly individual. Though forecasts of some sorts of event are public and shared with others, forecasts of other sorts of event are intrinsically private and personal. In particular, forecasts of how our personal relationships are likely to fare are not only of vastly more concern to
128
ourselves than to anyone else but are based on past experience and present information that are ours and ours alone. Thus, as regards the future, each one of us has his own personal forecasts of what good and what harm may befall. This is the private world of future expectations that each of us carries within. This theme is resumed in Chapter 14 in which attention is given especially to a person's forecasts of how his attachment figures are likely to behave and the immense influence these forecasts have on his propensity to be anxious or confident.
Cultural clues learnt from others
For long it has been suspected that children tend to 'catch' fears from their parents. Nevertheless, the extent to which there
-158-
is a correlation between what is feared by children and what by their parents is still little understood, and it is only during the past decade or so that the basic tendency to learn through observation has been the subject of systematic attention.
As a result of research it is now well established that learning through observation plays a significant part in the behavioural development of many species of bird and mammal ( Hinde 1970). In the case of humans, Bandura ( 1968), a leading exponent of social learning theory, claims that virtually anything that can be learnt through direct experience can be learnt vicariously through observing how others behave in particular situations and, especially, what the consequences of their behaviour are for them. In this way innumerable skills can be acquired. Observational learning provides a powerful means for the cultural transmission of which situations are to be avoided and which can be regarded as safe.
People concerned with children sometimes speak as though they thought it would be better were a child not to be influenced in what he fears by imitating 1 his parents. A moment's reflection, however, shows that on the contrary this is a wise provision of nature. Just as members of a band of non-human primates extend the range of the stimulus situations they avoid through imitating the behaviour of other animals (see Chapter 8), so do humans. Admittedly the consequence could on occasion be that some harmless situation was treated through several generations as though it were dangerous; yet more often, we may suppose, the tendency to imitate results in a young individual's being inducted quickly into the traditional wisdom of his social group and thereby avoiding hazards that might otherwise prove fatal.
Furthermore, to learn through imitating encompasses in the case of fear behaviour far more than learning to fear situations formerly not feared. It can equally well have an opposite effect. Thus the fear-arousing properties that a situation has for a child or an adult can be much reduced, or even extinguished, by his witnessing another person deal with the situation without fear and without harmful consequences. The restriction of the situations that arouse fear in an individual is discussed in Chapter 13.
____________________
1 There is a growing tendency in scientific literature to restrict the term 'imitation' to cases in
which a new motor pattern is developed. In what follows, however, the term is used in an everyday sense to denote that an individual observes the ways in which others respond to particular stimuli and then responds similarly, even though no new motor pattern is
129
involved. -159-
Reports of studies in which the degree to which situations feared by children are correlated with situations feared by their parents are unexpectedly scarce. Four can be quoted. In a study of seventy pre-school children, aged from two to six years, and their mothers, Hagman ( 1932) found significant correlations between the children who feared dogs and the mothers who feared dogs, and also between children and mothers who feared insects. A correlation was also present, though of lower degree, between children and mothers who feared thunderstorms. Not unexpectedly, when a child's fear of a situation was shared by his mother, the child was more likely to continue to be afraid of that situation than was a child whose mother was unafraid. In a comparable but better-controlled study, Bandura & Menlove ( 1968) also found a significant correlation between pre-school children who were afraid of dogs and parents (one or both) who were afraid of dogs. The third study concerns fear of dentistry. Shoben & Borland ( 1954) found that a most important factor in determining whether an individual will react with fear to the prospect of dental treatment is the attitude and experiences of members of his family. The fourth study concerns a hundred pre-school children evacuated with their mothers from a bombed area during the second world war. John ( 1941) reports a correlation of 0? 59 between the intensity of fear a child was reported to have shown during the raids and the intensity of fear his mother was reported to have shown. (Although the primary source of information was in most cases the mother herself, the existence of independent evidence for a few cases led the investigator to give the finding credence. )
Though the extent to which the situations that arouse fear tend to run in families and communities needs much further investigation, the ease with which fear of a previously neutral stimulus can be acquired vicariously is now well documented. For example, experiments in which the sound of a buzzer comes to arouse fear in a subject after he has observed that it is followed by an apparently painful shock to another person 1 are reported by Berger ( 1962) and also by Bandura & Rosenthal ( 1966). To observe another person apparently undergoing a shock whenever the buzzer sounds is found by many to be a most disagreeable experience. In Bandura's experiment some of the observers tried to decrease their discomfort by concen-
____________________
1 In these experiments the model who is observed is not in fact exposed to shock. Instead he
acts as though he has received a shock, e. g. by suddenly flexing his arm, dropping his pencil and wincing.
-160-
trating on other things. One remarked: 'When I noticed how painful the shock was to him I concentrated my vision on a spot which did not allow me to focus directly on either his face or hands. ' That the observers came also to respond to the stimulus with fear (as measured by their galvanic skin response) is therefore hardly surprising.
In the experimental situations described it is demanded that a subject observe what is going on. In real life we are free to observe or not as we wish. Though few systematic records are available, it seems probable that whenever we are in a strange or otherwise potentially
130
dangerous situation we commonly make a point of observing how others are responding and take our cue from them, especially when we believe them to be more experienced than we are ourselves. Children certainly do so. In the study referred to above (p. 160), Hagman ( 1932) conducted a series of simple experiments with his sample of pre-school children. He reports that, at the moment a feararousing stimulus was presented, nearly half the children looked tip at the adult who was with them. Schaffer ( 1971), it will be remembered, reports the same behaviour as early as twelve months (see p. 102 above).
Plainly this is a large area and very inadequately explored. It is also complex, since it is well known that the correlation between situations feared by children and those feared by adults is far from perfect. For example, a mother who is afraid of dogs and horses can have a daughter who is as bold as they come. Conversely, a father who is conspicuously unafraid can have a timid son. Many factors are evidently at work.
One point to which too little attention has so far been given in this exposition is that individuals learn to fear situations of certain types much more easily than they do others. This takes us back to the natural clues.
Continuing role of the natural clues
Already in this and the preceding chapter it has been emphasized that throughout our lives we tend to respond with fear to the natural clues -- to strangeness, sudden change of stimulation, rapid approach, height, being alone -- and to respond especially strongly to compound situations in which two or more natural clues are present together. Fear of animals and fear of darkness, both so common, are, it seems likely, to be explained by the fact that animals and darkness frequently constitute sources of two or more of the natural clues.
-161-
Fear of Animals
During the first eighteen months of life few children show fear of animals. Thenceforward, however, fear becomes increasingly easily aroused by animals so that during the third, fourth, and fifth years a majority of children are likely to show fear, at least on some occasions. Although thereafter the tendency for animals to arouse fear diminishes, it continues to be extremely common both in older children and in adults. (These findings were presented above, see Chapter 7).
On occasion, of course, a child may be threatened or even attacked by an animal, but it is unlikely that such events account for more than a very small proportion of the children who develop fear of animals. All the evidence suggests that in great part the readiness of children to develop such fear can be explained by the fact that animals are so frequently a source, simultaneously, of at least three of the natural clues that arouse fear, namely rapid approach, sudden movement, and sudden noise. This is well illustrated by an observation by Valentine ( 1930) who was among the first to study the ontogeny of human fear responses.
Valentine reports that one of his sons first showed fear of a dog at the age of twenty months. On this occasion a dog tripped over the string of the little boy's toy horse and yelped. In so doing it presented, of course, a combination of approach, sudden movement, by tripping, and sudden noise, by yelping. At this the little boy cried and thenceforward was afraid of the dog.
131
In that episode the dog belonged to a neighbour and so was presumably familiar to the child. On many other occasions, however, an animal behaving in this kind of way is a stranger. Whenever it is so, yet another natural clue is added to the constellation. Small wonder, then, that fear of animals is so widespread.
Not only do animals frequently present simultaneously several of the natural clues but there is reason to believe that they may present also some additional stimulus properties that increase the likelihood of a child's learning to fear them. Being furry is possibly one; a wriggling movement probably another; certain visual patterns may be still others.
Valentine had been struck by the ease with which young children seem to develop a fear of animals in comparison with fear of other things. He therefore conducted a little experiment
-162-
with one of his daughters, Y, 'an exceptionally healthy, strong and jovial youngster', aged at the time twelve and a half months ( Valentine 1930). In a first test the little girl, seated on her mother's knee, was given a pair of opera glasses which were then placed on a table in front of her. Each time she reached for them a wooden whistle was blown as loudly as possible behind her. Each time, she turned round quietly as if to see whence the noise had come. In these conditions the whistle aroused no fear. When the same afternoon, however, the same experiment was conducted, this time with a woolly caterpillar in place of the opera glasses, 'at once Y gave a loud scream and turned away from the caterpillar. This was repeated four times with precisely the same effect. ' Later the same day, while seated on her mother's knee and with no whistle being sounded, Y vacillated between showing interest in the caterpillar and turning away from it. When her brother picked up the leaf on which the caterpillar was crawling the little girl seemed to gain confidence and went to seize it (an example very probably of observational learning). From these experiments Valentine draws three conclusions. 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.
-140-
Whether suddenly alarmed or chronically anxious, whether temporarily comforted or steadily confident, the way a man or woman thinks and feels is determined in significant degree by these strong genetic biases to respond unthinkingly to the natural clues.
In the chapters that follow attention is given, first, to showing that the strong tendency to respond to the natural clues accounts for most of the more elaborate situations that humans come to fear, and, subsequently, to the way in which increasingly refined processes of appraisal lead to a broad spectrum of human feeling states. Before proceeding, however, let us consider further the genetically biased groundwork. We begin with the special place of physical pain as a natural clue.
The Limitations of Pain as a Natural Clue
In the past there have been theorists who have postulated that almost the only type of stimulus to which there is any genetic bias to respond with fear is physical pain, and that all other stimuli derive their fear-arousing properties from becoming associated with pain. Not only is the theory false, but a moment's thought shows it to be hardly plausible.
115
As a natural clue to potential danger, the experience of physical pain is in a special category. The clues to which attention has so far been directed are distal clues perceived by the distance receptors, eye, ear, and nose. By giving warnings while potential danger is still more or less remote, these clues enable an animal or man to take precautions in good time. By contrast, as noted in the previous chapter, to await events until pain is experienced may well be to wait too long. Whereas the distance receptors can be likened to advance look-outs, physical pain has the status of last ditch.
The special property of pain is that, being so late in acting, it leads to immediate and urgent action. The phase of alert wariness, so characteristic of many animals after a distal clue is first sensed, is absent. Instead, there is immediate and unthinking retreat, or, alternatively, attack.
Another special property of pain is, of course, its power to promote learning. Countless experiments demonstrate how rapidly and firmly an animal learns to recognize a situation in which it has experienced pain and to respond thenceforth by avoiding it. After such learning, an animal no longer relies on the hazardous proximal clue of pain but comes instead to use some distal clue that gives time and space in which it can take
-141-
precautions. The advance look-outs are alerted to identify and beware of a new clue.
Even though physical pain may be more highly correlated with potential danger than are some of the other natural clues, it is not infallible. For example, medical attention may be painful but is usually not dangerous; whereas a truly dangerous condition, such as internal haemorrhage, may be accompanied by no pain. That is but one example of a serious danger that is either without natural clues or heralded by faint ones only.
Dangers that have no Natural Clues
Earlier it was noted that the natural clues to which we react with fear are, singly and especially together, indicators of a high proportion of all the dangerous situations into which we might stray. Nevertheless, there are some dangerous situations that present no clue to which we have a natural bias to respond by escape. Some, indeed, even emit no signal that our sense organs can detect.
Among naturally occurring hazards, infectious illnesses are cases in point. Where infection is airborne there is usually no naturally occurring clue that we are able to sense and from which we are genetically biased to withdraw. (In contrast, by producing bad smells or tastes, food- and water-borne infections are often much less silent. ) In modern times, moreover, man has added a number of other dangers that also emit no clues to which human nature is sensitive. Examples are carbon monoxide gas and X-rays. Since in such cases evolution has as yet had neither time nor opportunity to provide us with natural means for their detection, we have to rely instead on man-made indicators.
Thus, although by exploiting the natural clues to danger and safety our genetic endowment provides us with a remarkably sensitive and efficient means of protection, it is far from foolproof. On countless occasions we are led unnecessarily to avoid wholly harmless situations; on a few others we are permitted to blunder into truly dangerous ones.
116
Potential danger of being alone
The natural clue to increased risk of danger with which this volume is especially concerned is being alone. Statistically, being alone is less safe than being with a companion. That that should be so during childhood, during sickness, and in old age
-142-
may not be too difficult to grasp. That it is also the case for the ordinary healthy grown-up man and woman may at first sight seem unexpected. Yet there is good reason to believe that it is so, especially in certain situations; even though in Western countries the situations may be few and the absolute risk not high. The thesis of this section is, therefore, that, as it was throughout man's earlier history, in many circumstances still today it is as appropriate to avoid being alone as it is to avoid any of the other natural clues to potential danger. That we should be so constructed that we find comfort in companionship and seek it, and that we experience greater or less degrees of anxiety when alone, is, therefore, in no way surprising.
In the previous volume (Chapter 4) it is argued that, if we are to understand human behavioural equipment, it is necessary to view it in the light of what we know of man's environment of evolutionary adaptedness. Later, following this line of thought (Chapter 12), the theory is advanced that in man's environment of evolutionary adaptedness the function of attachment behaviour, which of course promotes proximity to special companions, is protection from predators, and that this is as true for humans as it is for other species of mammal and bird. For all ground-living primates, safety lies in being with the band. To become separated from it is to provide a more or less easy meal for a lurking leopard 1 or a pack of hunting dogs. For weaker members, especially females and young, the old and the sick, isolation often spells speedy death.
By practical people this theory is sometimes treated as an academic curiosity. Yes, it may be said, there may well have been a time in man's history when separation entailed danger from predators. But that was long ago. For such responses to persist into modern times is an irrelevant nuisance. It is time to rid ourselves of such archaic superstition.
This line of reasoning has several defects. In the first place, even if we wished it, genetic biases built in over millions of years cannot be eradicated overnight. In the second, reflection suggests that to try to eradicate them might be most unwise. For in many parts of the world today the absolute risk attendant on
____________________
1 Since publication of the first volume further evidence has come to light of the dangers of
leopards to early man. According to Brain ( 1970) the fossilized bones of Paranthropus robustus found in a cave in the Transvaal are fragmented in ways typical of leopard prey. One of the betterpreserved skulls (of a juvenile) bears two holes the right size and distance apart to fit the canine teeth of a leopard.
-143-
being alone is still fairly high; and even in Western societies the risk may be higher than we like to imagine.
117
In Western countries today, it is true, injury and death are no longer due to predators. But there are other dangers. In place of predators, power-driven motor cars and household equipment provide novel hazards and take their toll. Young children newly mobile and the elderly are among the principal victims. Yet, though ordinary experience strongly suggests that those most at risk are children and old people left on their own, researchers concerned with accident prevention seem to have given the matter little attention. Statistics of traffic accidents for one of the London boroughs and also for Sweden are, however, revealing.
Traffic Accidents to Children
During 1968 in the London borough of Southwark, 1 injuries to pedestrians numbered 901, of which twenty-seven were fatal. Of the total injuries, 411, or nearly half (46 per cent), were to children under the age of fifteen. This represents an incidence of injury to children about three times that to adults.
The most vulnerable age-group were children between their fourth and eighth birthdays. At these ages the risk of injury was about five times that for adults. The incidence for those a little younger and a little older (the three- and the eight-year-olds) was only slightly lower. The age distribution was as shown in the table below.
Age in No. of
years
0- 2? 11 3- 5? 11 6- 8? 11 9-11? 11 12-14? 11 Total
injuries
14
125
124
81
67
411
Of all the children injured, almost two-thirds (62 per cent) were entirely alone. Even in the case of the younger children
____________________
1 I am indebted to Mr V. E. Golds, Road Safety Officer for the borough, for these figures.
-144-
more than half were alone. Many of the others were with other children, often no older than themselves. Only one in eight of the injured children was with an adult.
A similar picture emerges from Sweden ( Sandels 1971). The incidence of injury to pedestrians is especially high in children between their third and tenth birthdays. A special study of 177 accidents occurring to children under the age of eleven years at pedestrian
118
crossings shows that 44 per cent of the children were alone and another 34 per cent with peers; only one in five was with an adult.
From these figures we can conclude that the great excess of traffic accidents to children when compared with adults is due to their being out in the street either alone or with peers. 1 To anyone who has had the care of small children in an urban district this conclusion will hardly cause surprise.
Risk to Adults
It is perhaps easy to understand that for a young child or an old person to be alone is a risk. But, it may be protested, that can hardly be true also for a healthy adult. Reflection, however, strongly suggests that it is.
It seems very probable that, were comparative figures available, it would be found that even for healthy men and women in Western countries there are many situations in which risk of injury or death is greater when a person is alone than when in company. Walking in city streets at night is a case in point. It is not for nothing that in certain areas policemen patrol in pairs. Those who take part in active sports, moreover, are aware that to be alone carries added risk. Whether climbing mountains, swimming, exploring caves, or sailing the seas, to be alone is hazardous, sometimes because in detecting danger two heads are better than one, sometimes because an injury that would present no problem to a pair can prove fatal to a singleton.
____________________
1 Studies of the family background of children who are injured in traffic accidents ( Backett
& Johnston 1959; Burton 1968) cast light on why these children are not being looked after by a parent. When compared with children in a control group, more of the injured children are found to be unwanted and unloved and/or to have a mother who, currently, is anxiously preoccupied with other matters, e. g. illness in herself or in others in the household, younger siblings, elderly relatives, or her own pregnancy. Similar findings for children who sustain burns are reported by Martin ( 1970).
-145-
Yet another hazard to a man alone comes when he is overtaken by fatigue. Once asleep he cannot protect himself should danger threaten. When, by contrast, he has companions, each can take his turn on watch. The practice of alternating watches on board a ship at sea is, indeed, the organized and human version of a sleeping pattern common in birds that roost together in flocks and in primates that sleep together in bands. Because every animal is awake for some part of the night, at any one moment, while the majority of animals are asleep, a few are likely to be awake ready to give the alarm ( Washburn 1966).
It is true that in recent years great feats of single-handed navigation have been performed. That interest in their success should be so high is an earnest of the public's recognition, not only of the difficulties to be solved, but of the risks to be surmounted. Safety lies in numbers, especially in the companionship of familiars.
119
Potential safety of familiar companions and environment
Throughout these chapters it is emphasized that what is feared includes not only the presence, actual or imminent, of certain sorts of situation but the absence, actual or imminent, of certain other sorts of situation. Throughout life we tend to be drawn towards certain parts of the animate and inanimate environment, mainly people and places we are familiar with, and to be repelled by certain other parts of the environment, especially those that exhibit one or more of the natural clues to potential danger. Since two of the natural clues that tend to be avoided are strangeness and being alone, there is a marked tendency for humans, like animals of other species, to remain in a particular and familiar locale and in the company of particular and familiar people.
It has long been obvious that animals of any one species tend to restrict their movements so that they remain within those parts of the earth's surface to which they are physiologically adapted. Such parts can be defined in terms of various physical measures, such as earth, air or water, temperature gradients, rainfall, and also in terms of biological measures, such as presence or absence of certain foodstuffs. Only by regulating their movements in these ways are members of a species able to maintain the physiological measures on which life depends within certain critical limits. The types of behavioural system, activation and termination of which result in an animal's re-
-146-
maining within its ecological niche, are of the kind traditionally termed instinctive.
Yet, great though ecologically determined limitations may be, they are nothing in comparison with the limits constantly found in nature. It is still too little realized, perhaps, that the individuals of a species, so far from roaming at random throughout the whole area of the earth's surface ecologically suitable to them, usually spend the whole of their lives within an extremely restricted segment of it, known as the home range. 1 For example, a vole lives within its few hundred square yards of thicket, a troup of baboons within its dozen square miles of savanna, a band of human hunters and gatherers within its few hundred square miles of forest or plain. Even flocks of migrating birds, which may travel thousands of miles between nesting and wintering grounds, use only special parts of each: many birds nest each year at or very near the place they were born.
In a similar way birds and mammals do not mix indiscriminately with others of their kind. Individual recognition is the rule. With certain individuals close bonds may be maintained for long stretches of the life-cycle. With a number of others there may be a less close but none the less sustained relationship. Yet other individuals may either be of little interest or else be carefully avoided. Thus each individual has its own relatively small and very distinctive personal environment to which it is attached.
While the survival value of an animal's predisposition to remain within an ecologically suitable environment is clearly not in doubt, the survival value of its strong tendency to remain within its own special and familiar environment may at first sight seem debatable. Yet
120
examination of the issue shows that to do so in all likelihood confers distinct advantage, especially when conditions turn unfavourable. By remaining within a familiar environment an animal, or a human, knows at once where food and water are to be found, not only at different seasons of the ordinary year but also during those exceptionally bad years that occur from time to time; he knows, too, where shelter from the weather can be got, where there are trees or cliffs or caves
____________________
1 The concept of the home range embraces that of territoriality but is much broader. Whereas
very many species of bird and mammal show marked preferences for a particular home range (see Jewell & Loizos 1966), far fewer maintain and defend an exclusive territory. For a discussion of the probable functions of territory holding, which may differ between species, see Crook ( 1968).
-147-
that provide safety, what are the common dangers and from what quarter they are likely to come. By remaining in company with familiar companions, he can benefit from established and therefore relatively successful customs, for example food preferences, and also, when threatened by a predator, from coordinated social action. Thus, by remaining within his personal and familiar environment an individual stays within an arena that is comparatively safe and is kept well clear of many hazards that might otherwise endanger him.
Maintenance of an individual within his familiar environment is, it is postulated, the result of the activation and termination of behavioural systems that are sensitive to such stimulus situations as strangeness and familiarity, being alone and being with companions. On the one hand, behavioural systems mediating fear behaviour tend to remove the individual from situations that are potentially dangerous. On the other, those mediating attachment behaviour tend to lead him towards, or retain him within, situations that are potentially safe.
This brings the discussion back to attachment behaviour. The behavioural systems that maintain a younger or weaker individual in more or less close proximity to another discriminated and stronger individual can now be seen as a part of a larger set of systems that have the effect of so regulating the whole of a creature's movements that he remains as a rule within his familiar environment. Attachment to a parent figure is, in most species, ontogenetically the first form in which this type of behaviour develops.
In the next chapter and in Chapters 18 and 19 it is argued that many of the difficulties that have dogged psychiatric and psychoanalytic theories of anxiety have arisen because insufficient recognition has been given to the enormous roles that an individual's personal and familiar environment, including his familiar companions, plays in determining his emotional state. Only when it is realized that each man's environment is unique to himself can how he feels be understood.
Maintaining a stable relationship with the familiar environment: a form of homeostasis
Those trained in physiology may find it illuminating to view the behaviour under consideration as homeostatic. Whereas the systems studied by physiologists maintain certain
121
physicochemical measures, internal to the organism, within certain limits, the systems mediating attachment behaviour and fear
-148-
behaviour maintain the whole individual within a defined part of the environment. In the one case the states held steady are interior to the organism, in the other the states held steady concern the relationship of the organism to the environment.
A principal advantage of casting the theory of fear and anxiety presented here in terms of homeostasis is that by so doing it becomes possible to relate it to two other bodies of theory, both of which commonly invoke homeostatic principles. On the one hand, the theory can be linked to theories of stress and stress diseases, most of which invoke concepts of physiological homeostasis. On the other, it can be linked to a theory of defensive processes, which also, traditionally, have been conceived as contributing to the maintenance of some form of homeostasis. In place, however, of Freud's postulate that defensive processes help to reduce the level of stimulation in the mental apparatus to a steady low level, the theory advanced here conceives them as contributing to the maintenance of what can conveniently be termed a steady 'representational' state. 1
In the view presented here, maintenance of a steady relationship between an individual and his familiar environment occurs hardly less automatically and unthinkingly than does maintenance of his physiological steady states. In the case of each form of homeostasis, an individual is conceived as being born with a strong genetic bias to develop biological systems that, by being sensitive to certain types of stimuli, come into action whenever some particular measure deviates from certain set limits and cease action as soon as it is restored within those limits. Thus the theory proposed places the maintenance of a steady relationship between organism and familiar environment on a level of biological importance only one step lower than the maintenance of the much better understood physiological steady states.
Furthermore, the systems maintaining each of the two forms of homeostasis are seen as complementary. For it is clear that, so long as the systems that maintain an individual within his familiar environment are being successful, the loads placed on the systems that maintain physiological states steady are being eased. This is because, so long as an individual remains within his familiar physical environment and with familiar companions,
____________________
1 A theory of defensive processes incorporating these ideas is to be sketched in the later part
of the third volume, in which also a comparison will be made with certain other concepts of homeostasis that from time to time have been proposed by psychoanalysts.
-149-
he is more likely than he would be otherwise to find food and drink, and to achieve reliable and continuing protection from natural hazards -- from predators, from eating poisonous foodstuffs, from falling and drowning, from exposure and cold. Conversely, so long as the systems maintaining physiological homeostasis are successful, the healthier will the individual be and the easier will it be for him to maintain himself effectively within his familiar environment. Looked at in this light the regulatory systems that maintain a steady relationship between an individual and his familiar environment can be regarded as an 'outer
122
ring' of life-maintaining systems complementary to the 'inner ring' of systems that maintain physiological homeostasis.
It must, of course, be stressed that, whatever category of homeostasis is being considered, states are never maintained more than relatively stable nor, except rarely, do set-points and limits persist unchanged throughout the life-cycle. So long as the unit of study is the individual, indeed, processes of growth are the antithesis of processes of homeostasis. The principle of homeostasis is, therefore, only one among several. The reason for emphasizing it here is that homeostasis is held to be a key concept for understanding not only fear and anxiety but also grief and mourning.
While special attention is directed to the tendency for an individual to maintain a steady relationship between himself and his familiar environment, the important roles of exploratory and investigative behaviour, which tend to be antithetical to it, are not overlooked (see Volume I, Chapter 13). Nor are the developmental changes that occur in regular procession during the life-cycle.
-150-
Chapter 10
Natural Clues, Cultural Clues, and the Assessment of Danger
Innately, children seem to have little true realistic anxiety. . . . They will run along the brink of the water, climb on to the window-sill, play with sharp objects and with fire -- in short, do everything that is bound to damage them and to worry those in charge of them. When in the end realistic anxiety is awakened in them, that is wholly the result of education; for they cannot be allowed to make the instructive experiences themselves.
SIGMUND FREUD ( 1917b)
Clues of three kinds
Initially during infancy the only stimulus situations to which a child responds with fear are the natural clues. During the second and third years other situations are added, notably presence of animals, and darkness (with its associated happenings), both of which situations, it is argued, are readily learnt derivatives of the natural clues. From his second year, in addition, a child is much influenced by observing the behaviour of significant adults and imitating it. Among the great array of behaviour a child learns in this way, all of it culturally determined, is to respond with fear to a range of stimulus situations he has hitherto treated as neutral, or even interesting. These new stimulus situations are conveniently referred to as 'cultural clues'. In many instances, it is evident, imitative fear behaviour of this sort is entirely without insight into the nature of the danger being avoided. Because of this, fear aroused by a cultural clue has much in common with fear aroused by a natural clue. In neither case can the fear be regarded as 'realistic' in Freud's sense of the term.
Only very slowly and as his cognitive capacities develop does a child begin to distinguish natural or cultural clues from real danger and to learn methods of his own for calculating risk. During the same phase of growth all his behaviour is becoming organized increasingly in terms of goal-corrected plans, and
123
-151-
with it his fear behaviour. As a result of these linked developments his fear behaviour is becoming, as it is said, more 'rational' and 'realistic'. Thenceforward, during later childhood and adolescence, and into adult life, his capacity to assess real danger and to respond appropriately is likely steadily to improve.
Nevertheless, important though these new developments in the organization of fear behaviour may be, the bias to respond with fear to both the cultural and the natural clues persists. Indeed, not only throughout childhood but throughout adolescence and adult life as well the natural clues and their derivatives remain among the most effective of all the stimulus situations that arouse fear. Even the most courageous is not immune to fear on seeing some extraordinary apparition or a sudden rapid approach, or on hearing some piercing scream or finding himself alone in darkness in a strange place.
In intellectual circles both the persisting bias to respond to the natural clues and the value of that bias are all too often overlooked. As a result much human fear comes to be seen in a false perspective. For example, Arnold ( 1960), rightly impressed by the role of appraisal in the regulation of behaviour, goes so far as to assert that 'genuine fear develops only when the child is old enough to estimate the possibility of harm'. Elsewhere, in discussions of fear in human beings, there is commonly to be found the assumption, more or less explicit, that whereas fear of real danger is a healthy and often desirable response fear of anything else is childish or neurotic. Throughout psychiatry that assumption has for long been powerful and persuasive. It is found not only in the psychoanalytic tradition from Freud himself onwards (see above, Chapter 5) but in other traditions of psychiatry also (e. g. Lewis 1967). It is a main reason why the fear of separation from a loved figure is still so often, and so erroneously, held to be both neurotic and childish.
A principal thesis of this work is that the assumption that mature adults are afraid only of real danger, plausible though it may seem, is profoundly mistaken. Naturally enough an adult man, or woman, does what he, or she, can to calculate the prospects of real danger and to take the necessary precautions. Yet to make those calculations is often far from easy, and on some occasions would take dangerously long to do.
By contrast, to respond to the natural and cultural clues is quick and simple. To respond to the natural clues, moreover, especially when two or more people are present together, provides, as described
-152-
in the previous chapter, an efficient, if crude, system for minimizing danger and maximizing safety. No wonder, therefore, that alongside the more sophisticated measures for calculating danger, adult man continues to respond, at least tentatively, to each one of the natural clues and, when faced with compound situations, to respond especially strongly. Thus, in adult man fear behaviour comes to be elicited by clues that derive from at least three sources:
-- natural clues and their derivatives
-- cultural clues learnt by observation
-- clues that are learnt and used in more or less sophisticated ways in order to assess danger and avoid it.
Behaviour based on clues of the first type develops very early and is apt to be referred to as 'childish' and 'irrational'. Behaviour based on clues of the third type develops much later and
124
is commonly referred to as 'mature' and 'realistic'. Behaviour based on clues of the second type is intermediate: whether it is referred to as childish or mature, rational or irrational, turns on whether the onlooker shares or does not share the cultural norm reflected in the behaviour. For example, fear of ghosts is judged realistic by an observer from one culture and as childish by an observer from another.
A true evaluation of behaviour based on these three distinct types of clue is held to yield a picture very different from the popular one. Behaviour based on clues of the first and second types no less than behaviour based on those of the third is plainly consistent with normal development and mental health. In a healthily functioning individual, indeed, responses to clues of all three types are present; they can occur simultaneously or sequentially, and be either compatible with one another or in conflict.
In this chapter we consider the role of behaviour elicited by each of these three types of clue. Because so much attention has already been given to the natural clues we begin by considering the more sophisticated methods of assessing and avoiding danger.
Real danger: difficulties of assessment
Psychiatrists often speak as though it were easy to assess real danger. This is not so.
Both in ordinary life and in clinical practice there are two distinct types of problem. One is the difficulty that each of us has in assessing what is and what is not a real danger to his
-153-
own interests. Another is the difficulty that each of us has in assessing what is and what is not a real danger to another person.
Difficulties are met as soon as we try to define what we mean by 'real danger', whether to self or to another. There are a number of problems. One concerns how widely each of us draws the boundary of where his interests lie. A second concerns our understanding of what can, or cannot, cause injury. A third concerns the very varying ability of individuals to protect themselves and their interests: whereas a strong man might well be able to protect himself in a certain dangerous situation, a weaker one, or a woman or a child, might not.
We start with the problem of where each draws the boundary of his interests. Plainly, any situation that might lead to our own injury or death is classifiable as dangerous. The same would be agreed for anything that threatens injury or death to members of our family and to close friends. Beyond that, definition becomes more difficult. How widely do we extend the circle of friends and acquaintances whose safety we are concerned about? To what extent do we identify ourselves with the safety and wellbeing of the institution in which we work or the recreational club to which we belong? How do we rate threat to personal possessions, to house, and to favourite haunts?
Experience shows that a human being is constantly made afraid and anxious by threats of damage to a circle of persons, possessions, and places some way beyond himself and his body. For that reason it is necessary to include within the concept of real danger threat of injury or damage not only to the person himself but to the whole of his personal environment, as it is defined in the previous chapter.
125
All too often the need to include within the boundary the whole of an individual's personal environment is not recognized or, even if the principle is recognized, the nature and extent of the personal environment of a particular individual are not properly known. As a result, what is truly a danger for that person may go unnoticed by an onlooker.
Furthermore, not only is the nature of threat strictly relative to the person concerned but, as already remarked, means of protection are so too. Strong and competent people are able to protect themselves in situations in which weaker and less competent ones cannot.
Even when a definition of real danger is agreed, however, there remain great difficulties for each of us in assessing it.
-154-
For example, for an individual to calculate accurately when and in what degree he and his interests are endangered requires him to have a comprehensive knowledge of the world about him and to be able reliably to predict results. How many of us are qualified in these respects? It is easy to talk of real danger, but very difficult to estimate it.
It is indeed easy to forget that what is held to be publicly and permanently real is never more than some schematic representation of the world that happens to be favoured by a particular social group at a particular time in history. To some people during some periods to be afraid of ghosts is realistic. To other people during other periods to be afraid of germs is realistic. In matters of reality we all stand in danger of being arrogantly parochial.
That, however, is not to assert that everything is subjective, that there is no reality. The difficulty in using reality as a criterion lies, not in there being no reality, but in our imperfect capacity to comprehend it. That a child has an imperfect capacity to comprehend what is or may be truly dangerous is usually taken for granted. That the capacity of an adult is greater often by only a small margin tends to be forgotten.
To assess a risk of danger accurately requires us to take into account simultaneously a number of factors. Consider, for example, how we calculate the risk of being attacked by a particular dog. The ordinary dog, it will be agreed, is a harmless and amiable creature. Yet some dogs are dangerous to some people sometimes. What then are the criteria to be applied? On reflection we realize that they are numerous and complex. An accurate forecast rests partly on the sort of dog it is, partly on the situation in which we encounter it, partly on its behaviour and partly on how we estimate our own strength. Thus, we need to take into account the dog's age and sex, its breed, and perhaps also its probable training. Simultaneously, we should take into account whether the dog is on its home ground or elsewhere, whether with or without its master, and whether, in the case of a bitch, it has puppies. At the same time we should consider whether the dog is familiar with us, how it greets us, and how effective we judge ourselves to be in countering threat with threat and protecting ourselves should it attack. It is in fact a complex appraisal requiring considerable knowledge of dogs and accurate perception of the current situation. No wonder many adults as well as children despair of making it and behave as though all dogs are dangerous until
-155-
126
proved safe. Others, simplifying the situation in an opposite direction, may make the opposite assumption.
Consider again the difficulty of estimating accurately the danger of food-poisoning. To do so requires intimate knowledge of the food's origin, who has handled it, whether or not it has been cooked and the capabilities of various organisms to survive heating to various temperatures and for varying lengths of time. No wonder the ordinary housewife bases her behaviour on a limited number of culturally derived clues and practices.
In his capacity to assess and forecast real danger a child is of course even worse placed than an adult. Not only is he likely to be ill informed but, as Piaget has repeatedly shown (see Flavell 1963), his capacity to take into account more than a single factor at a time grows only slowly. It is fortunate that a child responds so readily to the natural and the cultural clues. Were he not to do so he would soon be dead.
'Imaginary' dangers
Assessment of danger always takes the form of a forecast. Sometimes the dangerous situation foreseen is judged to be imminent, at other times to be remote. In either case the likelihood of the situation's eventuating is of every degree. Dangerous situations that almost every adult in a society forecasts as 'probable present no problem. It is those situations that almost every adult forecasts as highly improbable or even impossible that are the challenge. Scoffingly, fear arising from such forecasts is dubbed 'exaggerated' or 'imaginary'; in more sober vein it is termed 'inappropriate'. For long, fear exhibited about such possibilities has constituted one of the principal riddles of psychopathology.
Yet, once the difficulty of making accurate forecasts of danger is grasped and once it is realized that if living beings are to survive there can be no great margin for error, the so-called imaginary fears come to be seen in a different and more sympathetic light. That children, still with a very imperfect model of the world, should at times gravely underestimate a danger may at times alarm us but is no surprise. That they should as frequently make an error of the opposite kind, foreseeing danger when we foresee none, is, when viewed in this perspective, no surprise either. Thus, when the bathwater goes down the plughole, how is a toddler to know he will not go down too? When, later, he hears tales of robbers and red indians intercepting coaches or robbing mail-trains, how is he
-156-
to know that he and his family may not be the next victims? The very great difficulty a child has in appraising at all accurately the degree of danger in which at any moment he may stand accounts, it is argued, for a much larger proportion of the so-called imaginary fears of childhood than is often supposed.
Sometimes 'imaginary' fear arises because of a simple misunderstanding, as when a small boy of six and a half, acting as a photographer's model, ran hurriedly off stage each time the photographer was about to press the button. Not until the next day was it revealed that, as soon as he heard the word 'shoot', he had run for his life. A similar type of misunderstanding led a boy of twelve, referred for stealing, to insist on having a sixpence in his pocket when he came to the clinic. The mystery was solved after some weeks when it transpired that he
127
believed the clinic to be a penal establishment, and that he had plans that if incarcerated he would escape, and he would then need the money for his bus fare home.
At other times 'imaginary' fear is a consequence of generalizing from too small a sample. If granny can die today, perhaps mother or father may die tomorrow. If a mother's first infant has died, is it surprising that she fears her second may die also?
The examples so far given are of erroneous or disproportionate forecasts of danger that arise from inaccurate or inadequate data. Until the source of an individual's erroneous forecast is known, his tendency to fear a particular situation will appear to another person to be absurd; it will also persist. Once the source is known about, however, the tendency is quickly seen by the other person to be far from unreasonable, even if misguided; and there is then a chance of its being corrected or modified.
In other cases fear of a situation that it may seem ridiculous to an outsider to fear can be explained in other ways. One source of such fear that has been greatly underestimated in the clinical literature is a forecast of danger that is in fact well based but remains inexplicable to an outsider because it is derived from information of the greatest secrecy. An example is seen in a child or adolescent one of whose parents is given to uttering dire threats -- of suicide, leaving home, even murder -- during emotional outbursts that, though real enough at the time, may be infrequent and in general out of character. While child or adolescent, not unnaturally, takes the threat seriously, the notion that such threats could ever be made may be discounted
-157-
or even denied by the parents. The key role that such family situations can play in accounting for the greatly intensified degree of separation anxiety suffered by some patients is considered in later chapters.
Another source of apparently unreasonable fear is a forecast of danger that derives from an individual's knowledge, conscious or unconscious, of certain desires of his own; for example, hostile wishes directed against someone he loves. Here again, to be afraid ceases to be unreasonable once the facts are known.
Yet other sources of fear that is or appears to be ill based lie in processes of projection and rationalization, to which brief attention is given in the next chapter.
In Chapters 18 and 19 further attention is given to some of the so-called irrational fears of anxious children and adults. These few paragraphs are intended to show only that the theoretical approach adopted can encompass without difficulty clinical problems of the greatest concern to every practising clinician and that a biological perspective in no way negates Freud's profoundly important discovery that fear can arise, not only from forecasts of how the external world and the people in it may behave, but also from forecasts of how we ourselves may possibly act.
Perhaps the most fundamental lesson to be learnt by anyone who wishes to understand the situations that other people fear is that forecasts of future dangers are as often as not strictly individual. Though forecasts of some sorts of event are public and shared with others, forecasts of other sorts of event are intrinsically private and personal. In particular, forecasts of how our personal relationships are likely to fare are not only of vastly more concern to
128
ourselves than to anyone else but are based on past experience and present information that are ours and ours alone. Thus, as regards the future, each one of us has his own personal forecasts of what good and what harm may befall. This is the private world of future expectations that each of us carries within. This theme is resumed in Chapter 14 in which attention is given especially to a person's forecasts of how his attachment figures are likely to behave and the immense influence these forecasts have on his propensity to be anxious or confident.
Cultural clues learnt from others
For long it has been suspected that children tend to 'catch' fears from their parents. Nevertheless, the extent to which there
-158-
is a correlation between what is feared by children and what by their parents is still little understood, and it is only during the past decade or so that the basic tendency to learn through observation has been the subject of systematic attention.
As a result of research it is now well established that learning through observation plays a significant part in the behavioural development of many species of bird and mammal ( Hinde 1970). In the case of humans, Bandura ( 1968), a leading exponent of social learning theory, claims that virtually anything that can be learnt through direct experience can be learnt vicariously through observing how others behave in particular situations and, especially, what the consequences of their behaviour are for them. In this way innumerable skills can be acquired. Observational learning provides a powerful means for the cultural transmission of which situations are to be avoided and which can be regarded as safe.
People concerned with children sometimes speak as though they thought it would be better were a child not to be influenced in what he fears by imitating 1 his parents. A moment's reflection, however, shows that on the contrary this is a wise provision of nature. Just as members of a band of non-human primates extend the range of the stimulus situations they avoid through imitating the behaviour of other animals (see Chapter 8), so do humans. Admittedly the consequence could on occasion be that some harmless situation was treated through several generations as though it were dangerous; yet more often, we may suppose, the tendency to imitate results in a young individual's being inducted quickly into the traditional wisdom of his social group and thereby avoiding hazards that might otherwise prove fatal.
Furthermore, to learn through imitating encompasses in the case of fear behaviour far more than learning to fear situations formerly not feared. It can equally well have an opposite effect. Thus the fear-arousing properties that a situation has for a child or an adult can be much reduced, or even extinguished, by his witnessing another person deal with the situation without fear and without harmful consequences. The restriction of the situations that arouse fear in an individual is discussed in Chapter 13.
____________________
1 There is a growing tendency in scientific literature to restrict the term 'imitation' to cases in
which a new motor pattern is developed. In what follows, however, the term is used in an everyday sense to denote that an individual observes the ways in which others respond to particular stimuli and then responds similarly, even though no new motor pattern is
129
involved. -159-
Reports of studies in which the degree to which situations feared by children are correlated with situations feared by their parents are unexpectedly scarce. Four can be quoted. In a study of seventy pre-school children, aged from two to six years, and their mothers, Hagman ( 1932) found significant correlations between the children who feared dogs and the mothers who feared dogs, and also between children and mothers who feared insects. A correlation was also present, though of lower degree, between children and mothers who feared thunderstorms. Not unexpectedly, when a child's fear of a situation was shared by his mother, the child was more likely to continue to be afraid of that situation than was a child whose mother was unafraid. In a comparable but better-controlled study, Bandura & Menlove ( 1968) also found a significant correlation between pre-school children who were afraid of dogs and parents (one or both) who were afraid of dogs. The third study concerns fear of dentistry. Shoben & Borland ( 1954) found that a most important factor in determining whether an individual will react with fear to the prospect of dental treatment is the attitude and experiences of members of his family. The fourth study concerns a hundred pre-school children evacuated with their mothers from a bombed area during the second world war. John ( 1941) reports a correlation of 0? 59 between the intensity of fear a child was reported to have shown during the raids and the intensity of fear his mother was reported to have shown. (Although the primary source of information was in most cases the mother herself, the existence of independent evidence for a few cases led the investigator to give the finding credence. )
Though the extent to which the situations that arouse fear tend to run in families and communities needs much further investigation, the ease with which fear of a previously neutral stimulus can be acquired vicariously is now well documented. For example, experiments in which the sound of a buzzer comes to arouse fear in a subject after he has observed that it is followed by an apparently painful shock to another person 1 are reported by Berger ( 1962) and also by Bandura & Rosenthal ( 1966). To observe another person apparently undergoing a shock whenever the buzzer sounds is found by many to be a most disagreeable experience. In Bandura's experiment some of the observers tried to decrease their discomfort by concen-
____________________
1 In these experiments the model who is observed is not in fact exposed to shock. Instead he
acts as though he has received a shock, e. g. by suddenly flexing his arm, dropping his pencil and wincing.
-160-
trating on other things. One remarked: 'When I noticed how painful the shock was to him I concentrated my vision on a spot which did not allow me to focus directly on either his face or hands. ' That the observers came also to respond to the stimulus with fear (as measured by their galvanic skin response) is therefore hardly surprising.
In the experimental situations described it is demanded that a subject observe what is going on. In real life we are free to observe or not as we wish. Though few systematic records are available, it seems probable that whenever we are in a strange or otherwise potentially
130
dangerous situation we commonly make a point of observing how others are responding and take our cue from them, especially when we believe them to be more experienced than we are ourselves. Children certainly do so. In the study referred to above (p. 160), Hagman ( 1932) conducted a series of simple experiments with his sample of pre-school children. He reports that, at the moment a feararousing stimulus was presented, nearly half the children looked tip at the adult who was with them. Schaffer ( 1971), it will be remembered, reports the same behaviour as early as twelve months (see p. 102 above).
Plainly this is a large area and very inadequately explored. It is also complex, since it is well known that the correlation between situations feared by children and those feared by adults is far from perfect. For example, a mother who is afraid of dogs and horses can have a daughter who is as bold as they come. Conversely, a father who is conspicuously unafraid can have a timid son. Many factors are evidently at work.
One point to which too little attention has so far been given in this exposition is that individuals learn to fear situations of certain types much more easily than they do others. This takes us back to the natural clues.
Continuing role of the natural clues
Already in this and the preceding chapter it has been emphasized that throughout our lives we tend to respond with fear to the natural clues -- to strangeness, sudden change of stimulation, rapid approach, height, being alone -- and to respond especially strongly to compound situations in which two or more natural clues are present together. Fear of animals and fear of darkness, both so common, are, it seems likely, to be explained by the fact that animals and darkness frequently constitute sources of two or more of the natural clues.
-161-
Fear of Animals
During the first eighteen months of life few children show fear of animals. Thenceforward, however, fear becomes increasingly easily aroused by animals so that during the third, fourth, and fifth years a majority of children are likely to show fear, at least on some occasions. Although thereafter the tendency for animals to arouse fear diminishes, it continues to be extremely common both in older children and in adults. (These findings were presented above, see Chapter 7).
On occasion, of course, a child may be threatened or even attacked by an animal, but it is unlikely that such events account for more than a very small proportion of the children who develop fear of animals. All the evidence suggests that in great part the readiness of children to develop such fear can be explained by the fact that animals are so frequently a source, simultaneously, of at least three of the natural clues that arouse fear, namely rapid approach, sudden movement, and sudden noise. This is well illustrated by an observation by Valentine ( 1930) who was among the first to study the ontogeny of human fear responses.
Valentine reports that one of his sons first showed fear of a dog at the age of twenty months. On this occasion a dog tripped over the string of the little boy's toy horse and yelped. In so doing it presented, of course, a combination of approach, sudden movement, by tripping, and sudden noise, by yelping. At this the little boy cried and thenceforward was afraid of the dog.
131
In that episode the dog belonged to a neighbour and so was presumably familiar to the child. On many other occasions, however, an animal behaving in this kind of way is a stranger. Whenever it is so, yet another natural clue is added to the constellation. Small wonder, then, that fear of animals is so widespread.
Not only do animals frequently present simultaneously several of the natural clues but there is reason to believe that they may present also some additional stimulus properties that increase the likelihood of a child's learning to fear them. Being furry is possibly one; a wriggling movement probably another; certain visual patterns may be still others.
Valentine had been struck by the ease with which young children seem to develop a fear of animals in comparison with fear of other things. He therefore conducted a little experiment
-162-
with one of his daughters, Y, 'an exceptionally healthy, strong and jovial youngster', aged at the time twelve and a half months ( Valentine 1930). In a first test the little girl, seated on her mother's knee, was given a pair of opera glasses which were then placed on a table in front of her. Each time she reached for them a wooden whistle was blown as loudly as possible behind her. Each time, she turned round quietly as if to see whence the noise had come. In these conditions the whistle aroused no fear. When the same afternoon, however, the same experiment was conducted, this time with a woolly caterpillar in place of the opera glasses, 'at once Y gave a loud scream and turned away from the caterpillar. This was repeated four times with precisely the same effect. ' Later the same day, while seated on her mother's knee and with no whistle being sounded, Y vacillated between showing interest in the caterpillar and turning away from it. When her brother picked up the leaf on which the caterpillar was crawling the little girl seemed to gain confidence and went to seize it (an example very probably of observational learning). From these experiments Valentine draws three conclusions. 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.
