The behaviour of the infants during episode 5, after mother had returned, is
referred
to again in Chapter 21.
Bowlby - Separation
Another three, namely Freud's earlier theory of transformed libido ( 1905b) and both of Klein's theories, of persecutory and depressive anxiety ( 1934; 1935), had different origins and came only later to be applied to the problem of separation anxiety.
All these five theories, however, are distinctly complex since in each case the author rules out of court the idea that absence of mother could, in and of itself, be the real cause of the distress and anxiety seen.
Consequently each author either feels constrained to search for a reason of some other kind or else applies a theory developed in another context.
Only occasionally has a student of the problem accepted the data at their face value and presented a theory of a sixth type, one that regards the distress and
35
subsequent anxiety as primary responses not reducible to other terms and due simply to the nature of a child's attachment to his mother. Among those who have advanced this view are Suttie ( 1935), Hermann ( 1936), and, with some qualifications, Fairbairn ( 1943; 1963) and Winnicott (e. g. 1952). Half a century earlier, it is interesting to note, William James ( 1890) had recorded his view that 'The great source of terror in infancy is solitude'.
-31-
The very similar generalization made by Freud in 1905, and quoted at the head of the next chapter, shows that he was early aware of the data. Indeed, as Strachey makes apparent in his introduction to the Standard Edition of Inhibitions, Symptoms and Anxiety ( SE 20: 77 - 86 ), from that time onwards the anxiety manifested by a young child when separated from his mother is constantly in Freud's mind and he returns to it repeatedly whenever he makes a further attempt to solve the problem of anxiety. Nevertheless, because the basic postulates from which he began his theorizing biased him in other directions, Freud himself never adopted a theory of the sixth type.
These varied attempts to account for the phenomena of separation anxiety are not only of historical interest but of great practical importance, because each theory gives rise to a different model of personality functioning and psychopathology and, in consequence, to significantly different ways of practising psychotherapy and preventive psychiatry. Because of their continuing and living influence, a detailed review of psychoanalytic theories of separation anxiety is presented in Appendix I. Some of the assumptions on which they rest are evaluated in Chapter 5 in the light of present knowledge of biology and ethology.
Before discussing theory further, however, it is useful to consider additional observations of behaviour during and after separation, starting with the behaviour of human children and proceeding thence to a comparison with the behaviour of young of other species. In all the studies to be described, it must be emphasized, either mother leaves child or child is removed more or less unwillingly from mother. The very different behaviour seen in the reverse situation, in which mother remains in a known place while child explores, is described already in the first volume (Chapter 13) and is the subject of papers by Anderson ( 1972a, b, c) and Rheingold & Eckerman ( 1970). Provided a child initiates the movement himself, in the certain knowledge of where mother can be found, he is not only content but often adventurous. In what follows the word 'separation' always implies that the initiative is taken either by mother or by some third party.
-32-
36
Chapter 3
Behaviour with and without Mother: Humans
Anxiety in children is originally nothing other than an expression of the fact that they are feeling the loss of the person they love.
SIGMUND FREUD ( 1905b)
Naturalistic observations
Chapter I presented data concerning the behaviour of young children when they are away from home and placed for days or weeks either in a residential nursery or in a foster home. Here, by contrast, we are concerned with separation situations of much shorter duration. We begin with separations lasting from a day to a few hours, all requiring a child to be in a strange place with strange people and without substitute mothering.
A number of psychologists have made records of the behaviour of young children when they first enter nursery school or go to a research centre for examination. In so doing the psychologists have, usually without intending it, amassed evidence that to start nursery school much before the third birthday is for most children an undesirably stressful experience. The records, indeed, make it apparent that ignorance of the natural history of attachment behaviour, coupled with a misguided enthusiasm that small children should quickly become independent and 'mature', has resulted in practices that expose children, and their parents, to a great deal of unnecessary anxiety and distress. Nevertheless, for scientific purposes the resulting records have the great advantage that there is no danger that the degree of upset has been exaggerated; indeed the reverse is probably the case.
The first and largest study of this kind seems to have been undertaken by Shirley at the Harvard School of Public Health ( Shirley & Poyntz 1941; Shirley 1942). In this study, 199 children (101 boys and 98 girls) between the ages of two and eight years were observed in the course of an all-day visit to a research centre, during they which were subjected to a variety of
-33-
psychological and medical examinations, interspersed with periods for play, meals, and rest. The children were without mother throughout the day. The authors express the belief that 'the responses of these children to separation from their mothers were fairly typical for children who are cared for predominantly by their mothers during their pre-school period'.
All children visited the centre at six-monthly intervals; and any one child attended over a period of about three years. Age at starting varied: twenty-five paid their first visit at two years, a further twenty-eight at two and a half, and other batches at each half-year of age up to about five. As a result the number of children observed at each age-level varied from twenty- five at the age of two years to a maximum of 127 at the age of five and a half years. Results are given in terms of the percentage of children of each age-level and sex who were upset in
37
each of three situations met with during the day -- leaving mother, play period at centre, and meeting mother at end of day. Published data do not distinguish between the responses of children on their first visit to the centre and those of children who had paid one or more previous visits.
Of the children aged from two to four years, about half 1 are reported as upset on leaving mother at the day's start and half also on meeting mother again at the end of the day. The proportion upset diminishes in the older age-groups, though at the day's end it never falls below 30 per cent in the case of the boys. Even during the free play period with a congenial though strange mother-surrogate the percentage of children upset is substantial, varying from about 40 per cent of the youngest children (two to three years old) to about 20 per cent at four years and 15 per cent of the older ones (five- to seven-year-olds):
. . . emotional upsets during the play period were by no means uncommon. The children manifested their uneasiness in the playroom in a variety of ways in addition to crying and calling for mother. Some merely stood disconsolately killing time; some shifted uneasily from one foot to the other; some peered out the window disappointedly searching for father's familiar car in the stream of traffic. . . . Such children ignored the proffered toys and resisted the play suggestions that were offered.
____________________
1 Because results are expressed in percentages and the N varies for each age-group and sex,
it is not possible to calculate an exact percentage for larger categories of children. -34-
Some sat distracted, aimlessly fiddling with a toy or sifting sand. Of the younger children, half were explicit in expressing a desire for mother; of those between four and a half and six years of age, the proportion asking for mother dropped to about one-quarter.
A number of children who had not expressed upset during the day's proceedings showed it on reunion with mother:
Usually it was the child who had bravely winked back the tears and made a determined effort to surmount his feelings of insecurity earlier in the day that gave way to his pent-up emotion in tears. At the sight of mother his needs for autonomy and independence vanished, and he reverted to the degree of babyishness he had overcome early in the morning.
At each age-level, proportionally fewer girls than boys were overtly upset. Moreover, when girls were upset, the intensity and duration of upset were less than in boys. It is evident that the authors of the study approve the girls for their greater 'maturity' and regret the 'babyishness' of the boys.
The authors note that the three-year-olds tended to be more upset than both the younger children and those older: 'Children of two and two and a half years were little aware of what the day would bring forth; they had little anticipatory dread. ' By three years they were 'more aware of the demands of the day, and more reluctant to leave home'. This was true especially of those who had paid one or two previous visits to the centre. Naturally, both the physical examination and the psychological testing were carried out as kindly as possible. Nevertheless, in mother's absence, so far from becoming used to the six-monthly
38
examinations, the children became more apprehensive about them: 'Familiarity with the situation from one or two previous experiences seemed to make the children grow more apprehensive' and they tended to be more upset when the day began ( Shirley 1942). By contrast, children aged five years and upwards were more likely to settle down, and some are reported to have enjoyed the day.
A study that derives from Shirley's but is confined to a very small part of the area covered by the earlier investigation is described by Heathers ( 1954). Not only is the age-range restricted to the youngest children but the behaviour reported is confined to the way in which they respond to being removed from home to go to a nursery school.
Thirty-one children between the ages of twenty-three and
-35-
thirty-seven months, from middle-class homes and of above average intelligence, were observed during their first five days of starting nursery school. On each of those days each child was called for by a student he had not previously met and taken off to school by car. To meet the researcher's requirements, each 'child's mother was asked to part with him at the door and to let the observer take him out to the car'. Although each mother had attempted to explain to her child what was in store for him, it is very doubtful whether explanation could have conveyed much to such young children. Observed behaviour was checked against a list of eighteen items, which conveys a vivid picture of the kinds of response likely to occur. It runs as follows: When taken from home to car
1. Cries
2. Hides, tries to hide, etc.
3. Resists getting dressed to go
4. Clings to mother
5. Calls for mother
6. Tries to go back to house
7. Must be carried to car
8. Resists being carried to car
During first five minutes in car
9. Cries
10. Calls for mother
11. Seeks reassurance or comforting
12. Resists reassurance or comforting
13. Tense, withdrawn or unresponsive
When arrives at school and enters building
14. Cries
15. Resists leaving car
16. Must be lifted out, carried
17. Clings to trip observer
18. Holds back, reluctant to enter.
The daily scores of the thirty-one children on the above eighteen items ranged during the five days from zero to 13, and thus show great individual variation. On the first day the mean score was 4? 4. Although by the fifth day the upset score of twenty-one children was lower than it had been on the first day, in the case of four children it was higher.
39
It is of interest to note that, on the first day, the older
-36-
children (aged from thirty to thirty-seven months) were significantly more upset than the younger ones (twenty-three to twenty-nine months); on succeeding days, however, there was no difference between children of the two age-groups. Heathers follows Shirley in noting the possibility that the slightly older children were more upset at first because, having experienced more previous visits to the research centre for purposes of testing, they were more apt to foresee what was going to happen.
A third study in the same tradition as those of Shirley and Heathers is already referred to in the first volume of this work (Chapter 11) where a brief account is given of observations made by Murphy ( 1962) of children visiting a research centre for a planned play session. In this later study the arrangements for collecting the children by car resembled those adopted by Heathers, but parting from mother was handled very differently. Though the children were encouraged to go off in the car on their own with an escort, the escort was not entirely strange. Furthermore, no obstacle was put in the way of mother going too, should child protest or mother prefer to accompany him. It is no surprise that only a small minority of the fifteen children aged between two and a half and four years agreed to go without mother. On arriving at the centre, however, mother departed leaving the child alone.
Murphy's findings are consistent with those of the earlier studies. Her records of individual children include some that give a clear account of a child's determination to have his mother accompany him. There is good reason to believe that such a reaction is entirely healthy and natural for a young child in a situation in which he is being invited to accompany two ladies he hardly knows to an unknown destination.
A detailed descriptive study by Janis ( 1964) of one little girl, who began attending nursery school for two half-days a week when she was no more than two years and three months old, illustrates well both how anxious a child of this age is made by the experience and how such anxiety can be hidden, at least for a time.
Lottie is described as 'a normal, highly verbal child', the youngest of three girls in a professional family. Her parents are described as 'sensitive to the needs of their children [and] aware of the possibility of separation difficulties'. The nursery school itself had a policy whereby a child's mother stayed with him at school until he seemed ready to be there alone.
During the first two occasions that Lottie attended, mother
-37-
stayed with her. On the third occasion, when mother left her briefly, Lottie, laughing, called out repeatedly 'Mommy! Daddy! Dorrie! Heidi! ' (the names of her two sisters, aged respectively five and three-quarters and ten and a half). A week later, on the fifth visit, Lottie insisted on wearing a skirt like one worn by Dorrie, to whom she was strongly attached. By the fourteenth session Lottie was claiming to be Dorrie: 'I'm Dorrie. Call me Dorrie. '
During the next few sessions, however, Lottie began objecting more strongly than she had done earlier to her mother leaving, and she occasionally cried for her. On the day before the
40
eighteenth session Lottie (at home) insisted on following her mother around the house and holding on to her. On the next day, at nursery school, ' Lottie bursts into tears when her mother says good-bye. She cried hard . . . her face hot and flushed. ' Thenceforward, Lottie ceased to call herself Dorrie.
Instead of the steady improvement that her mother anticipated with Lottie's acknowledgement of nursery school as her own, Lottie's behaviour deteriorates. She cannot let her mother leave at all; she cries bitterly when her mother does go; she clings more and more to her mother in school, less able to play independently than before; her play is limited, regressed, uncontrolled and violent at times; she loses urinary control at home, in token fashion (a few minor accidents), for the first time after being completely dry for half a year.
Furthermore, during these weeks, whenever Lottie was left at home with a familiar person while her mother went out, she showed increasingly intense longing for her mother. She also became increasingly obstinate and disobedient.
During the early sessions of the next school term, which started when Lottie was two years and six months, Lottie insisted on her mother staying with her. Later, though she accepted that her mother should leave, she was listless and halfhearted in her play; and on mother's return Lottie's first remark was: 'I didn't cry. ' By the end of four weeks, however, she was once again crying when her mother left, and this continued on and off for the rest of the term. In the upshot, it was not until the third term, which started when Lottie was two years and nine months, that she began to settle happily at school without her mother. 1
____________________
1 Lottie's methods of coping in mother's absence, for example by claiming to be a big girl
like her sister, will be discussed in Volume III. -38-
Although Lottie's parents are described as sensitive to their children's needs and the nursery school re? gime as benign, it is evident from the account that both of Lottie's parents and the teacher were expecting far too much of so young a child. Much pressure, it is clear, was put on her not to cry. Although often she succeeded in controlling herself, her constant preoccupation with not crying, which runs through the report, is evidence of the strain she was under.
Had there not been so many misconceptions about the norms of behaviour to be expected of young children when left, even briefly, in a strange place with strange people, it would have been unnecessary to present these data so fully. Yet misconceptions persist, especially among professional people. Again and again it is implied that a healthy normal child should not make a fuss when mother leaves, and that if he does so it is an indication either that mother spoils him or that he is suffering from some pathological anxiety. It is hoped that such reactions will be seen in a new and more realistic light when the natural history and function of attachment behaviour are understood.
Experimental studies
Because subjecting a child to a very brief separation, lasting only a few minutes, is ethically permissible, the behaviour to which it gives rise can be examined in experimental conditions;
41
variables can therefore be controlled and detailed systematic observation is relatively easy. Moreover, the behaviour of a child when his mother is absent can be compared with his behaviour when she is present, with other conditions remaining unchanged.
The first to undertake such studies was Arsenian ( 1943). In recent years a number of other workers have followed suit, for example Ainsworth ( Ainsworth & Wittig 1969; Ainsworth & Bell 1970), Rheingold ( 1969), Cox & Campbell ( 1968), Maccoby & Feldman ( 1972), Lee, Wright & Herbert (in preparation), and Marvin ( 1972). The overall picture that emerges of the behaviour of children as it develops from the first birthday through to the third is a consistent one.
Ainsworth's study is alluded to briefly in the first volume of this work (Chapter 16) when patterns of attachment are under consideration. Since that volume was written, she and her colleagues have published observations in more detail and on a much larger sample of children (for a recent review of findings see Ainsworth, Bell & Stayton, in press).
-39-
The subjects of Ainsworth's study are fifty-six infants, aged one year, of white American middle-class families, reared within their family in ways typical of the 1960s. In respect of a subsample of twenty-three infants, detailed observations were made throughout the first year of life of the development of their social behaviour, with particular reference to attachment behaviour. In respect of the other thirty-three infants, limited observations of development were begun during their ninth month ( Bell 1970). Then the behaviour of all fifty-six babies was observed around the time of the first birthday. When a baby neared his birthday, 1 his mother was invited to participate with him in a brief series of experimental episodes, the purpose of which was to learn how the infant would behave in a congenial though slightly strange setting, first in the presence of mother and later in her absence.
To this end Ainsworth furnished a small room with three chairs and an open space in the middle. A chair near one end of the room was for mother, another, at the same end and opposite, was for a stranger, and a small chair at the other end had toys heaped upon it. The situation was designed to be novel enough to excite a child's interest but not so strange that he would be frightened. The entry of the stranger (female) was intended to be so gradual that any fear it evoked could be attributed to her unfamiliarity and not to any abrupt or alarming behaviour. There were eight experimental episodes, arranged so that the least disturbing ones came first; moreover, the series as a whole was similar in kind to many that an infant might be expected to encounter in his ordinary life. Both mother and stranger were instructed in advance regarding the roles they were to play. The episodes were arranged as follows.
In a preliminary episode a mother, accompanied by one of the observers, carried her infant into the room; the observer then left.
During episode 2, which lasted three minutes, mother put her infant down between the two chairs meant for the adults and then sat quietly in her own chair. She was not to participate in her infant's play unless he sought her attention, and then to do so only a little.
At the start of episode 3, which also lasted three minutes, the stranger entered. For one minute she sat quietly in her chair; then, for a second minute, she conversed with mother; finally,
42
____________________
1 Thirty-three of the babies were aged forty-nine and fifty weeks, twenty-three were fifty-
one weeks old. -40-
for a third minute, she gently approached the infant showing him a toy. Meanwhile mother sat quietly.
Episode 4 began with mother leaving the room unobtrusively, leaving her handbag on the chair. If the infant was playing happily the stranger stayed quiet; but if he was inactive she tried to interest him in a toy. Should an infant become distressed she did what she could to distract him or comfort him. Like the previous two episodes, this one lasted three minutes; but if an infant was much distressed and could not be comforted the episode was curtailed.
Episode 5 began with mother's return, after which the stranger departed. On entering, mother was to pause in the doorway in order to see what her infant's spontaneous response to her return would be. Thenceforward she was free to do whatever suited -- to comfort him if required and to settle him afresh in play with the toys. Once he was settled she was to leave the room again, pausing briefly as she went to say 'bye-bye'.
During episode 6, in consequence, the infant was left all alone. Unless curtailed because of distress, this episode lasted the usual three minutes.
Thereafter first stranger and then mother returned to make episodes 7 and 8.
Throughout the series of episodes the behaviour of infant, mother, and stranger was recorded by observers from behind a one-way vision window. From the narrative record two measures of behaviour could be obtained for each infant: (a) the frequency with which different sorts of behaviour were shown during each episode, the frequency in every case being measured by scoring 1 for each period of fifteen seconds during which that behaviour was seen (thus, for a three-minute episode a score could range from zero to 12); (b) the intensity of certain kinds of behaviour shown during each episode; in making ratings of intensity it was often necessary to take account of how mother or stranger was behaving to the infant.
The finding to which attention is specially drawn in this chapter is that the behaviour of these fifty-six one-year-old infants during the episodes when mother was absent (nos. 4 and 6) was in every case much changed from what it had been during the earlier episode (no. 2) when mother was sitting quietly in the room with them. All the infants showed behaviour of a kind that everyone would describe as anxious or distressed, and as being due to his missing his mother.
During episode 2 while his mother was present, the typical
-41-
picture of an infant was one of active interest in the scene. As a rule he moved around freely and played with the toys, giving only an occasional glance towards his mother; a small minority (seven infants), however, were inactive and tended to stay where they had been put.
43
During this episode crying was conspicuous by its absence, though an occasional child whimpered for a few moments to begin with.
During episode 3, in which the stranger joined mother and infant, the behaviour of most of the children changed substantially. Staring at the stranger was almost universal; many infants moved rather closer to mother; and exploration and play diminished, on average to about half what they had been. Some infants showed a tendency to cry or grizzle; but in only five cases was crying of any intensity. As a rule the stranger was treated with interest and, fairly soon, with cautious friendliness.
During episode 4 mother had departed and infant found himself alone with stranger. Half the infants showed a strong tendency to seek mother, starting usually as soon as they realized she had gone. Eleven followed mother to the door or struggled to do so; the others either looked at the door frequently or for long periods, or else searched for mother in the chair on which she had been sitting. There was also much crying and other signs of distress. For the group as a whole there was four times as much crying during mother's absence as there had been during episode 3. A dozen infants cried practically the whole time and another thirteen for a part of it. In all, thirty-nine infants either cried or searched, or did both (thirteen cases). This leaves seventeen, a fairly substantial minority, who did neither.
The behaviour of the infants during episode 5, after mother had returned, is referred to again in Chapter 21. Suffice it to say that half of them actively approached mother and showed a clear desire to be close to her, while another six either signalled or approached in a less purposeful way. Thirteen of the more active ones, having achieved close physical contact with mother, maintained it both by clinging to her and by resisting her attempts to put them down. All who had been crying stopped doing so although some infants who had been acutely distressed were not comforted quickly.
During episode 6, after mother had departed again, this time leaving her infant all alone, searching and crying were seen in more infants and were also more intense than they had been during episode 4. On this occasion forty-four infants searched for mother, of whom thirty-one followed her to the door. Of the
-42-
thirty-one, fourteen banged on the door or tried to open it: they either reached vainly for the knob or tried to insert their fingers into the crack. Among the dozen infants who did not search for mother, there were a few who had searched for her during the first separation episode but who on the second occasion showed only distress. During episode 6 there was also a great deal of crying. Forty infants cried more or less strongly; these included all those who had been distressed during the earlier episode as well as many others. Some rocked, or kicked their heels on the floor, or moved at random 'like a little trapped animal'. Only two infants neither searched nor cried; thirty did both.
Episode 6 was ended by stranger's return and the start of episode 7. After three minutes, during which infant was with stranger, mother returned and episode 8 began.
During episode 8, the tendencies to approach mother, to cling to her, and to resist being put down were much stronger and were seen in more of the infants than had been the case during the previous reunion episode. This time thirty-five of the fifty-six infants actively approached
44
mother and showed an evident desire for physical contact; a further nine either signalled their desire for contact or else approached mother in a less purposeful way. Two other infants, though they did not approach mother, engaged in lively interchange with her across a distance. Most striking was the large number of infants (forty-two) who both actively clung to mother and resisted being put down by her; another three, though they did not cling, resisted being put down.
A minority of infants included in the above figures showed signs of ambivalence towards mother of greater or less degree. Thus a few were seen to ignore mother briefly before approaching her, and others alternated between approaching and turning away. There were a few others who were so ambivalent that they mingled active attempts to seek and maintain contact with mother with attempts to get away from her.
A further small minority of infants (seven) behaved quite differently: they neither approached mother nor showed any desire to do so. Instead, they persistently ignored her and refused to respond when she invited them to come. Some even avoided looking at her.
To return to episodes 4 and 6: when behaviour shown during these two episodes, when mother was absent, is examined it is found that an infant was extremely likely either to search or to
-43-
cry or to do both together. The table below shows the number of children who responded in one of these three ways for each episode.
Behaviour
Cry only Search only Search and cry Total (N = 56)
Episode
4 6
12 10 14 14 13 30 39 54
When the behaviour of any one infant in episode 6 is compared with what he showed in episode 4 the following sequences are found:
-- those who cried only in episode 4 were likely to do the same in episode 6
-- those who searched only in episode 4 were likely to search and cry in episode 6
-- those who searched and cried in episode 4 were likely to do the same in episode 6, though a few cried only.
The individual differences in the responses of these children are of great interest and (as is discussed in Chapter 21) are correlated with the different patterns of mother-child interaction observed in the preceding year. Here, however, our concern is with the features that the infants' responses had in common. On each occasion when mother left the room, first leaving infant with stranger and then leaving him all alone, the behaviour of every infant changed. Play and exploratory behaviour either slowed down or ceased altogether. On the second
45
occasion especially, all but two of the infants showed marked dislike of the situation and expressed it by seeking mother, by crying unhappily, or by doing both together. The extent of distress and anxiety during mother's absence was considerable, even though the room and the toys remained exactly as they had been.
While to describe the broad features of behaviour for the sample as a whole enables generalizations to be framed with confidence, it tends also to be a little impersonal. To illustrate something of what the series of episodes meant to one small boy
-44-
and his mother, a description is given of one case, selected to be as representative as any one case can be: 1
1. Mother, Baby, Observer. Brian had one arm hooked over his mother's shoulder as they
came into the room; he was holding on to her, grasping a fold of her blouse. He looked
around soberly, but with interest, at the toys and at the observer.
2. Mother, Baby. After being put down, Brian immediately crept towards the toys and
began to explore them. He was very active, picking toys up, then dropping them or moving them about, with vigorous movements. He crept around quite a bit, mostly on his mother's side of the room. Although his attention was fixed on the playthings, he glanced up at his mother six times, and smiled at her twice. She glanced at him covertly, from time to time, but their glances did not seem to meet. Once he threw a toy with a clatter at her feet; she moved it back towards him. Otherwise there was no interaction between them. Towards the end of the three minutes he blew into a long cardboard tube, vocalizing as though pretending it were a horn, and then he looked up at his mother with a smile, seeming to expect her to acknowledge his accomplishment.
3. Stranger, Mother, Baby. He turned to look at the stranger when she entered, with a pleasant expression on his face. He played with the tube again, vocalized, smiled, and turned to glance at his mother. He continued to play, glancing at the stranger twice. When the stranger and his mother began to converse, he continued to explore actively at the end of the room, and looked up only once -- at the stranger. Towards the end of this minute of conversation he crept over to his mother, pulled himself up, and stood briefly, holding on to her knee with one hand, and clutching her blouse with the other. Then he turned back to play. When the stranger began her approach by leaning forward to offer him a toy he smiled, crept towards her, and reached for it. He put the toy in his mouth. She offered him the tube and he blew into it again. He looked back and forth from the toys to the stranger and did not look at his mother at all.
4. Stranger, Baby. He did not notice his mother leave. He continued to watch the stranger and the toys she was manipulating. Suddenly, he crept to his mother's chair, pulled
____________________
1 This description is taken from Ainsworth & Wittig ( 1969).
-45-
himself up into a standing position, and looked at the stranger. She tried to distract him with a pull-toy. He approached the toy, and began to roll it back and forth; but he glanced again at his mother's empty chair. He was less active than he had been when alone with his mother, and after two minutes his activity ceased. He sat chewing the string of the pull-toy, and glancing from the stranger to his mother's chair. He made an unhappy noise, then a cry-face, and then he cried. The stranger tried to distract him by offering him a block; he took it but
46
then threw it away with a petulant gesture. He gave several more little protesting cries, but he did not cry hard.
5. Mother, Baby. When his mother opened the door and paused in the doorway, Brian looked at her immediately and vocalized loudly, with a quality that could have been either a laugh or a cry; then he crept to her quickly, and pulled himself up, with her help, to hold on to her knees. Then she picked him up, and he immediately put his arms around her neck, his face against her shoulder, and hugged her hard. He then gave her another big hug before she put him down. He resisted being put down; he tried to cling to her and protested loudly. Once on the floor, he threw himself down, hid his face in the rug, and cried angrily. His mother knelt beside him and tried to interest him in the toys again. He stopped crying and watched. After a moment she disengaged herself and got up to sit on her chair. He immediately threw himself down and cried again. She helped him to stand, and cuddled him. For a moment he reciprocated in the cuddle, but then he threw himself down on the floor again, crying. She again picked him up, and tried, to direct his attention to a squeaky ball. He looked at it, still holding on to his mother, with one arm hooked over her shoulder. He began to play, but quickly turned back to his mother with a brief cry, and clung to her. This alternation of play and clinging continued. After four and a half minutes, his mother, apparently not wishing to delay us, picked a moment when he was interested in a ball, and moved to the door.
6. Baby Alone. As she said 'bye-bye' and waved, Brian looked up with a little smile, but he shifted into a cry before she had quite closed the door. He sat crying, rocking himself back and forth. He cried hard, but occasionally lulled a little and looked around. After a minute and a half the episode was curtailed and stranger instructed to enter.
-46-
7. Stranger, Baby. Brian lulled slightly when he saw the stranger enter, but he continued to cry. She first tried to distract him, then offered her arms to him. Brian responded by raising his arms; she picked him up, and he stopped crying immediately. She held him in her arms, and showed him the pictures tacked up around the edges of the mirrorwindow. He looked with apparent interest; he held on to her tightly, grasping a fold of her clothing. Occasionally he gave a little sob, but for the most part he did not cry. But when she put him down, he screamed. She picked him up again, and he lulled.
8. Mother, Baby. At the moment that his mother returned Brian was crying listlessly. He did not notice his mother. The stranger half-turned and pointed her out. Brian looked towards her, still crying, and then turned away. But he soon 'did a double take'. He looked back and vocalized a little protest. His mother offered her arms to him. He reached towards her, smiling, and leaned way out of the stranger's arms and his mother took him. He threw his arms around her neck, hugging her hard, and wiggling with excitement. Then the stranger tried to attract his attention. Brian did not notice her advance until she touched him; he immediately clung to his mother and buried his face in her shoulder. His mother continued to hold him, and he cuddled and clung to her, as the episode ended.
Since Ainsworth first reported her findings the results of several other studies have come to hand. In three of them ( Maccoby & Feldman 1972; Marvin 1972) the series of experimental situations used was planned to be as similar to Ainsworth's as possible, but in each the children were older. In a further two studies ( Cox & Campbell 1968; Leeet al. , in preparation) the situations differ from those used by Ainsworth, but in each there is opportunity to study children in an experimental setting first with mother present and later with mother absent. Details are given in the table overleaf.
47
Because the last two studies listed give data comparing behaviour at one year with that to be seen at two years and onwards it is convenient to present findings from one of them first.
Preliminary findings reported by Lee and his colleagues show that, while attachment behaviour continues to be extremely active at the time of the second birthday, the behavioural system
-47-
Authors
Maccoby & Feldman Maccoby & Feldman Marvin
Cox & Campbell
Leeet al.
Ages at children studied
2, 2 1/2, 3 years 2 1/2 years
2, 3, 4 years
14 months 24-37 months 1, 2, 3 years
which
Samples
White American 30-60, longitudinal
From Israeli kibbutzim 20, cross-sectional
White American 3 * 16, cross-sectional
White Canadian
2 * 20, cross-sectional
Middle-class English 27, longitudinal
governing it has altered in many respects since the first birthday. A comparison of the behaviour of the same children placed in the same situation at one year and two years of age shows that, at two years of age, children are likely:
-- to maintain greater proximity to mother -- a finding already reported from observations made out of doors by Anderson ( 1972a)
-- to be more hesitant in approaching a stranger.
On the other hand, merely to be close to mother and to be able to see her seem sufficient to give a child of two years a sense of security, whereas a one-year-old is likely to insist on physical contact. Two-year-olds, moreover, protest less than do oneyear-olds during a brief period when mother leaves them alone. Lee concludes that, compared with one-year-olds, two-yearolds have available more sophisticated cognitive strategies for maintaining contact with mother. They make much more use of looking and verbal communication, and are probably also using mental imagery in ways hardly possible for a one-yearold. As a result their attachment behaviour is better organized and their proximity-keeping is more proficient than it was when they were a year younger.
During the third year of life changes in the behaviour seen in the experimental situations are probably also in large part a result of the developments occurring in a child's cognitive competence. In their longitudinal study of children between
48
-48-
their second and third birthdays, Maccoby & Feldman ( 1972) note the much greater ability of three-year-olds to communicate with mother over a distance and also their increased ability to understand, when mother leaves the room, that she will soon be returning. As a result, when the reaction of three-year-olds to mother's brief absence is compared with that of two-year- olds, both crying and going to the closed door show a marked decrease. In addition, three- year-olds who have been left alone recover their equanimity when they are rejoined even by a stranger, whereas two-year-olds remain as upset when the stranger returns as they were when left entirely alone.
The responses of the children observed by Maccoby & Feldman when tested at the intermediate age of two and a half years in the same series of situations were roughly intermediate between the responses seen at two years and those seen at three years. Interestingly enough, the behaviour shown in the same situations by kibbutzim children at the age of two and a half differed very little from that of the American children at that age. Similarities between the groups were found in regard both to the means for the groups and to the range of individual variation within them. These findings are in keeping with other observations that suggest that the development of attachment behaviour in children brought up in kibbutzim is in most cases very similar to that of children brought up in traditional families (see Volume I, Chapter 15).
Although attachment behaviour develops in important ways during the second and third years of life, a child's behaviour in these experimental situations when mother is absent continues to be very different from what it is when she is present. For example, Maccoby & Feldman found that the manipulative play of two-year-olds decreased by about one-quarter when they were left with a stranger and by about half when they were left alone. Conversely, the proportion of children who cried increased enormously, from 5 per cent when mother was present to 30 per cent when the children were left with a stranger and to 53 per cent when they were left alone. Changes in the behaviour of three-year-olds when mother was absent were less striking than in two-year-olds but in the same direction. Their manipulative play decreased by one-sixth when they were left with a stranger and by one-third when they were left alone. The proportion of children who cried rose from zero to 5 per cent and 20 per cent respectively in the two situations in which mother was absent.
-49-
In addition to crying after mother had departed, there were many children at each age-level who showed a desire to follow her. Of the two-year-olds, 30 per cent not only went to the door but made efforts to open it; and a further 21 per cent stood near the door or leaned against it. Of the three-year-olds, 34 per cent attempted to open the door, and nearly half of them banged on it vigorously. At each age-level, again, a substantial minority of children expressed anger at mother's absence: 19 per cent at two years, 31 per cent at two and a half years, and 14 per cent at three years.
In noting the increased activity observed when children were left alone, especially evident at the ages of two and two and a half, Maccoby & Feldman write:
This increased activity frequently took the form of anxious searching or agitated movement. There was occasionally a quite opposite kind of reaction to the stress of being alone: a kind of
49
frozen immobility. . . . some children stood very still. This might occur near the door, when the child appeared to be waiting for his mother's return, or it might occur elsewhere in the room. In a few instances the child played with the toys, but each movement occurred at a markedly reduced speed, much as though the action had been rendered on a slow-motion film. Also, it occasionally happened that a child who was upset over separation would alternate between an unfocused running activity and immobility.
As regards the evaluation of these observations it is perhaps necessary to remind readers that on each occasion when mother departed she was absent for no more than three minutes, and for an even shorter time if the child was distressed, and that on the first of the two occasions the child was left with a friendly female stranger whom he had first met in the presence of mother. Furthermore, the toys he had been playing with were still there.
The findings of a cross-sectional study of samples of eight boys and eight girls at each of three age-levels by Marvin ( 1972) are in broad agreement with those of Maccoby & Feldman and here observations are extended to the fourth birthday. In Marvin's study the behaviour of boys and girls tended to differ. The two-year-old boys were as much upset as were Ainsworth's one-year-olds. Three-year-old boys were less upset than twoyear-olds; and four- year-old boys were comparatively little
-50-
affected by any of the situations. By contrast, the two- and three-year-old girls were appreciably less affected by events than were one-year-olds, whereas the four-year-old girls were much more upset, especially by being left alone. An explanation suggested by Marvin of this last result, which is unexpected, is that a four-year-old girl may be especially disturbed by mother's apparently arbitrary behaviour in the test situation and by her unwillingness to cooperate when the child asks her not to leave her alone. Although in broad outline the findings of these different studies are consistent, there are many differences of detail. For example, neither Ainsworth with her one-year-olds nor Maccoby & Feldman with their two- and three-year-olds found sex differences of any magnitude; whereas Lee and his colleagues with their one- and two-year-olds and also Marvin with his twos, threes, and fours were struck by the differences between boys and girls. This and other discrepancies in the results reported in different studies are not easy to interpret. It seems not unlikely that relatively small differences in the arrangements for the testing, for example, in the behaviour of the stranger, can affect considerably the intensity, though not the form, of any behaviour exhibited. From these and other miniature separation experiments certain conclusions can be drawn:
a. In a benign but slightly strange situation, young children aged between eleven and thirty- six months, and brought up in families, are quick to notice mother's absence and commonly show some measure of concern, varying considerably but amounting very often to obvious, and in some cases to intense, anxiety and distress. Play activity decreases abruptly and may cease. Efforts to reach mother are common.
b. A child of two years is likely to be almost as upset in these situations as a child of one, and at neither age is he likely to make a quick recovery when rejoined either by mother or by a stranger.
c. A child of three is less likely to be upset in these situations and is more able to understand that mother will soon return. On being rejoined by mother or a stranger he is relatively quick to recover.
d. A child of four may either be little affected by the situations or else be much distressed by mother's apparently arbitrary behaviour.
50
-51-
e. As children get older they are able to use vision and verbal communication as means for keeping in contact with mother; should they become upset when mother leaves the room older children will make more determined attempts to open the door in order to find her.
f. Up to 30 per cent of children are made angry by mother's leaving them alone in these circumstances.
g. In some studies and at some ages no differences are observed in the behaviour of boys and girls. In so far as any differences are observed, boys tend to explore more in mother's presence and to be more vigorous in their attempts to reach her when she has gone; girls tend to keep closer to mother and also to make friends more readily with the stranger.
A further finding from these miniature separation experiments, and one that links with the findings of Shirley ( 1942) and Heathers ( 1954) (see pp. 35 and 37 above), is only very recently reported. This is that when a child of about one year is tested in Ainsworth's series of episodes for a second time, a few weeks after the first testing, he is more upset and anxious than he was on the first occasion. When mother is present he keeps closer to her and clings more tightly. When she is absent he cries more ( Ainsworth, personal communication). These findings emerge from a test--retest study of twenty-four babies tested first at fifty weeks of age and a second time two weeks later. On the assumption that increased sensitivity is not due simply to maturation, which is unlikely, these findings provide the first experimental evidence that at one year of age a separation lasting only a few minutes, in what would ordinarily be regarded as a bland situation, is apt to leave a child more sensitive than he was before to a repetition of the experience.
Ontogeny of responses to separation The First Year
Since the responses to separation that are so unmistakable in infants of twelve months and older are not present at birth, it is clear that they must develop at some time during the first year of life. Unfortunately, studies designed to throw light on this development are few, and are confined to infants admitted to hospital.
35
subsequent anxiety as primary responses not reducible to other terms and due simply to the nature of a child's attachment to his mother. Among those who have advanced this view are Suttie ( 1935), Hermann ( 1936), and, with some qualifications, Fairbairn ( 1943; 1963) and Winnicott (e. g. 1952). Half a century earlier, it is interesting to note, William James ( 1890) had recorded his view that 'The great source of terror in infancy is solitude'.
-31-
The very similar generalization made by Freud in 1905, and quoted at the head of the next chapter, shows that he was early aware of the data. Indeed, as Strachey makes apparent in his introduction to the Standard Edition of Inhibitions, Symptoms and Anxiety ( SE 20: 77 - 86 ), from that time onwards the anxiety manifested by a young child when separated from his mother is constantly in Freud's mind and he returns to it repeatedly whenever he makes a further attempt to solve the problem of anxiety. Nevertheless, because the basic postulates from which he began his theorizing biased him in other directions, Freud himself never adopted a theory of the sixth type.
These varied attempts to account for the phenomena of separation anxiety are not only of historical interest but of great practical importance, because each theory gives rise to a different model of personality functioning and psychopathology and, in consequence, to significantly different ways of practising psychotherapy and preventive psychiatry. Because of their continuing and living influence, a detailed review of psychoanalytic theories of separation anxiety is presented in Appendix I. Some of the assumptions on which they rest are evaluated in Chapter 5 in the light of present knowledge of biology and ethology.
Before discussing theory further, however, it is useful to consider additional observations of behaviour during and after separation, starting with the behaviour of human children and proceeding thence to a comparison with the behaviour of young of other species. In all the studies to be described, it must be emphasized, either mother leaves child or child is removed more or less unwillingly from mother. The very different behaviour seen in the reverse situation, in which mother remains in a known place while child explores, is described already in the first volume (Chapter 13) and is the subject of papers by Anderson ( 1972a, b, c) and Rheingold & Eckerman ( 1970). Provided a child initiates the movement himself, in the certain knowledge of where mother can be found, he is not only content but often adventurous. In what follows the word 'separation' always implies that the initiative is taken either by mother or by some third party.
-32-
36
Chapter 3
Behaviour with and without Mother: Humans
Anxiety in children is originally nothing other than an expression of the fact that they are feeling the loss of the person they love.
SIGMUND FREUD ( 1905b)
Naturalistic observations
Chapter I presented data concerning the behaviour of young children when they are away from home and placed for days or weeks either in a residential nursery or in a foster home. Here, by contrast, we are concerned with separation situations of much shorter duration. We begin with separations lasting from a day to a few hours, all requiring a child to be in a strange place with strange people and without substitute mothering.
A number of psychologists have made records of the behaviour of young children when they first enter nursery school or go to a research centre for examination. In so doing the psychologists have, usually without intending it, amassed evidence that to start nursery school much before the third birthday is for most children an undesirably stressful experience. The records, indeed, make it apparent that ignorance of the natural history of attachment behaviour, coupled with a misguided enthusiasm that small children should quickly become independent and 'mature', has resulted in practices that expose children, and their parents, to a great deal of unnecessary anxiety and distress. Nevertheless, for scientific purposes the resulting records have the great advantage that there is no danger that the degree of upset has been exaggerated; indeed the reverse is probably the case.
The first and largest study of this kind seems to have been undertaken by Shirley at the Harvard School of Public Health ( Shirley & Poyntz 1941; Shirley 1942). In this study, 199 children (101 boys and 98 girls) between the ages of two and eight years were observed in the course of an all-day visit to a research centre, during they which were subjected to a variety of
-33-
psychological and medical examinations, interspersed with periods for play, meals, and rest. The children were without mother throughout the day. The authors express the belief that 'the responses of these children to separation from their mothers were fairly typical for children who are cared for predominantly by their mothers during their pre-school period'.
All children visited the centre at six-monthly intervals; and any one child attended over a period of about three years. Age at starting varied: twenty-five paid their first visit at two years, a further twenty-eight at two and a half, and other batches at each half-year of age up to about five. As a result the number of children observed at each age-level varied from twenty- five at the age of two years to a maximum of 127 at the age of five and a half years. Results are given in terms of the percentage of children of each age-level and sex who were upset in
37
each of three situations met with during the day -- leaving mother, play period at centre, and meeting mother at end of day. Published data do not distinguish between the responses of children on their first visit to the centre and those of children who had paid one or more previous visits.
Of the children aged from two to four years, about half 1 are reported as upset on leaving mother at the day's start and half also on meeting mother again at the end of the day. The proportion upset diminishes in the older age-groups, though at the day's end it never falls below 30 per cent in the case of the boys. Even during the free play period with a congenial though strange mother-surrogate the percentage of children upset is substantial, varying from about 40 per cent of the youngest children (two to three years old) to about 20 per cent at four years and 15 per cent of the older ones (five- to seven-year-olds):
. . . emotional upsets during the play period were by no means uncommon. The children manifested their uneasiness in the playroom in a variety of ways in addition to crying and calling for mother. Some merely stood disconsolately killing time; some shifted uneasily from one foot to the other; some peered out the window disappointedly searching for father's familiar car in the stream of traffic. . . . Such children ignored the proffered toys and resisted the play suggestions that were offered.
____________________
1 Because results are expressed in percentages and the N varies for each age-group and sex,
it is not possible to calculate an exact percentage for larger categories of children. -34-
Some sat distracted, aimlessly fiddling with a toy or sifting sand. Of the younger children, half were explicit in expressing a desire for mother; of those between four and a half and six years of age, the proportion asking for mother dropped to about one-quarter.
A number of children who had not expressed upset during the day's proceedings showed it on reunion with mother:
Usually it was the child who had bravely winked back the tears and made a determined effort to surmount his feelings of insecurity earlier in the day that gave way to his pent-up emotion in tears. At the sight of mother his needs for autonomy and independence vanished, and he reverted to the degree of babyishness he had overcome early in the morning.
At each age-level, proportionally fewer girls than boys were overtly upset. Moreover, when girls were upset, the intensity and duration of upset were less than in boys. It is evident that the authors of the study approve the girls for their greater 'maturity' and regret the 'babyishness' of the boys.
The authors note that the three-year-olds tended to be more upset than both the younger children and those older: 'Children of two and two and a half years were little aware of what the day would bring forth; they had little anticipatory dread. ' By three years they were 'more aware of the demands of the day, and more reluctant to leave home'. This was true especially of those who had paid one or two previous visits to the centre. Naturally, both the physical examination and the psychological testing were carried out as kindly as possible. Nevertheless, in mother's absence, so far from becoming used to the six-monthly
38
examinations, the children became more apprehensive about them: 'Familiarity with the situation from one or two previous experiences seemed to make the children grow more apprehensive' and they tended to be more upset when the day began ( Shirley 1942). By contrast, children aged five years and upwards were more likely to settle down, and some are reported to have enjoyed the day.
A study that derives from Shirley's but is confined to a very small part of the area covered by the earlier investigation is described by Heathers ( 1954). Not only is the age-range restricted to the youngest children but the behaviour reported is confined to the way in which they respond to being removed from home to go to a nursery school.
Thirty-one children between the ages of twenty-three and
-35-
thirty-seven months, from middle-class homes and of above average intelligence, were observed during their first five days of starting nursery school. On each of those days each child was called for by a student he had not previously met and taken off to school by car. To meet the researcher's requirements, each 'child's mother was asked to part with him at the door and to let the observer take him out to the car'. Although each mother had attempted to explain to her child what was in store for him, it is very doubtful whether explanation could have conveyed much to such young children. Observed behaviour was checked against a list of eighteen items, which conveys a vivid picture of the kinds of response likely to occur. It runs as follows: When taken from home to car
1. Cries
2. Hides, tries to hide, etc.
3. Resists getting dressed to go
4. Clings to mother
5. Calls for mother
6. Tries to go back to house
7. Must be carried to car
8. Resists being carried to car
During first five minutes in car
9. Cries
10. Calls for mother
11. Seeks reassurance or comforting
12. Resists reassurance or comforting
13. Tense, withdrawn or unresponsive
When arrives at school and enters building
14. Cries
15. Resists leaving car
16. Must be lifted out, carried
17. Clings to trip observer
18. Holds back, reluctant to enter.
The daily scores of the thirty-one children on the above eighteen items ranged during the five days from zero to 13, and thus show great individual variation. On the first day the mean score was 4? 4. Although by the fifth day the upset score of twenty-one children was lower than it had been on the first day, in the case of four children it was higher.
39
It is of interest to note that, on the first day, the older
-36-
children (aged from thirty to thirty-seven months) were significantly more upset than the younger ones (twenty-three to twenty-nine months); on succeeding days, however, there was no difference between children of the two age-groups. Heathers follows Shirley in noting the possibility that the slightly older children were more upset at first because, having experienced more previous visits to the research centre for purposes of testing, they were more apt to foresee what was going to happen.
A third study in the same tradition as those of Shirley and Heathers is already referred to in the first volume of this work (Chapter 11) where a brief account is given of observations made by Murphy ( 1962) of children visiting a research centre for a planned play session. In this later study the arrangements for collecting the children by car resembled those adopted by Heathers, but parting from mother was handled very differently. Though the children were encouraged to go off in the car on their own with an escort, the escort was not entirely strange. Furthermore, no obstacle was put in the way of mother going too, should child protest or mother prefer to accompany him. It is no surprise that only a small minority of the fifteen children aged between two and a half and four years agreed to go without mother. On arriving at the centre, however, mother departed leaving the child alone.
Murphy's findings are consistent with those of the earlier studies. Her records of individual children include some that give a clear account of a child's determination to have his mother accompany him. There is good reason to believe that such a reaction is entirely healthy and natural for a young child in a situation in which he is being invited to accompany two ladies he hardly knows to an unknown destination.
A detailed descriptive study by Janis ( 1964) of one little girl, who began attending nursery school for two half-days a week when she was no more than two years and three months old, illustrates well both how anxious a child of this age is made by the experience and how such anxiety can be hidden, at least for a time.
Lottie is described as 'a normal, highly verbal child', the youngest of three girls in a professional family. Her parents are described as 'sensitive to the needs of their children [and] aware of the possibility of separation difficulties'. The nursery school itself had a policy whereby a child's mother stayed with him at school until he seemed ready to be there alone.
During the first two occasions that Lottie attended, mother
-37-
stayed with her. On the third occasion, when mother left her briefly, Lottie, laughing, called out repeatedly 'Mommy! Daddy! Dorrie! Heidi! ' (the names of her two sisters, aged respectively five and three-quarters and ten and a half). A week later, on the fifth visit, Lottie insisted on wearing a skirt like one worn by Dorrie, to whom she was strongly attached. By the fourteenth session Lottie was claiming to be Dorrie: 'I'm Dorrie. Call me Dorrie. '
During the next few sessions, however, Lottie began objecting more strongly than she had done earlier to her mother leaving, and she occasionally cried for her. On the day before the
40
eighteenth session Lottie (at home) insisted on following her mother around the house and holding on to her. On the next day, at nursery school, ' Lottie bursts into tears when her mother says good-bye. She cried hard . . . her face hot and flushed. ' Thenceforward, Lottie ceased to call herself Dorrie.
Instead of the steady improvement that her mother anticipated with Lottie's acknowledgement of nursery school as her own, Lottie's behaviour deteriorates. She cannot let her mother leave at all; she cries bitterly when her mother does go; she clings more and more to her mother in school, less able to play independently than before; her play is limited, regressed, uncontrolled and violent at times; she loses urinary control at home, in token fashion (a few minor accidents), for the first time after being completely dry for half a year.
Furthermore, during these weeks, whenever Lottie was left at home with a familiar person while her mother went out, she showed increasingly intense longing for her mother. She also became increasingly obstinate and disobedient.
During the early sessions of the next school term, which started when Lottie was two years and six months, Lottie insisted on her mother staying with her. Later, though she accepted that her mother should leave, she was listless and halfhearted in her play; and on mother's return Lottie's first remark was: 'I didn't cry. ' By the end of four weeks, however, she was once again crying when her mother left, and this continued on and off for the rest of the term. In the upshot, it was not until the third term, which started when Lottie was two years and nine months, that she began to settle happily at school without her mother. 1
____________________
1 Lottie's methods of coping in mother's absence, for example by claiming to be a big girl
like her sister, will be discussed in Volume III. -38-
Although Lottie's parents are described as sensitive to their children's needs and the nursery school re? gime as benign, it is evident from the account that both of Lottie's parents and the teacher were expecting far too much of so young a child. Much pressure, it is clear, was put on her not to cry. Although often she succeeded in controlling herself, her constant preoccupation with not crying, which runs through the report, is evidence of the strain she was under.
Had there not been so many misconceptions about the norms of behaviour to be expected of young children when left, even briefly, in a strange place with strange people, it would have been unnecessary to present these data so fully. Yet misconceptions persist, especially among professional people. Again and again it is implied that a healthy normal child should not make a fuss when mother leaves, and that if he does so it is an indication either that mother spoils him or that he is suffering from some pathological anxiety. It is hoped that such reactions will be seen in a new and more realistic light when the natural history and function of attachment behaviour are understood.
Experimental studies
Because subjecting a child to a very brief separation, lasting only a few minutes, is ethically permissible, the behaviour to which it gives rise can be examined in experimental conditions;
41
variables can therefore be controlled and detailed systematic observation is relatively easy. Moreover, the behaviour of a child when his mother is absent can be compared with his behaviour when she is present, with other conditions remaining unchanged.
The first to undertake such studies was Arsenian ( 1943). In recent years a number of other workers have followed suit, for example Ainsworth ( Ainsworth & Wittig 1969; Ainsworth & Bell 1970), Rheingold ( 1969), Cox & Campbell ( 1968), Maccoby & Feldman ( 1972), Lee, Wright & Herbert (in preparation), and Marvin ( 1972). The overall picture that emerges of the behaviour of children as it develops from the first birthday through to the third is a consistent one.
Ainsworth's study is alluded to briefly in the first volume of this work (Chapter 16) when patterns of attachment are under consideration. Since that volume was written, she and her colleagues have published observations in more detail and on a much larger sample of children (for a recent review of findings see Ainsworth, Bell & Stayton, in press).
-39-
The subjects of Ainsworth's study are fifty-six infants, aged one year, of white American middle-class families, reared within their family in ways typical of the 1960s. In respect of a subsample of twenty-three infants, detailed observations were made throughout the first year of life of the development of their social behaviour, with particular reference to attachment behaviour. In respect of the other thirty-three infants, limited observations of development were begun during their ninth month ( Bell 1970). Then the behaviour of all fifty-six babies was observed around the time of the first birthday. When a baby neared his birthday, 1 his mother was invited to participate with him in a brief series of experimental episodes, the purpose of which was to learn how the infant would behave in a congenial though slightly strange setting, first in the presence of mother and later in her absence.
To this end Ainsworth furnished a small room with three chairs and an open space in the middle. A chair near one end of the room was for mother, another, at the same end and opposite, was for a stranger, and a small chair at the other end had toys heaped upon it. The situation was designed to be novel enough to excite a child's interest but not so strange that he would be frightened. The entry of the stranger (female) was intended to be so gradual that any fear it evoked could be attributed to her unfamiliarity and not to any abrupt or alarming behaviour. There were eight experimental episodes, arranged so that the least disturbing ones came first; moreover, the series as a whole was similar in kind to many that an infant might be expected to encounter in his ordinary life. Both mother and stranger were instructed in advance regarding the roles they were to play. The episodes were arranged as follows.
In a preliminary episode a mother, accompanied by one of the observers, carried her infant into the room; the observer then left.
During episode 2, which lasted three minutes, mother put her infant down between the two chairs meant for the adults and then sat quietly in her own chair. She was not to participate in her infant's play unless he sought her attention, and then to do so only a little.
At the start of episode 3, which also lasted three minutes, the stranger entered. For one minute she sat quietly in her chair; then, for a second minute, she conversed with mother; finally,
42
____________________
1 Thirty-three of the babies were aged forty-nine and fifty weeks, twenty-three were fifty-
one weeks old. -40-
for a third minute, she gently approached the infant showing him a toy. Meanwhile mother sat quietly.
Episode 4 began with mother leaving the room unobtrusively, leaving her handbag on the chair. If the infant was playing happily the stranger stayed quiet; but if he was inactive she tried to interest him in a toy. Should an infant become distressed she did what she could to distract him or comfort him. Like the previous two episodes, this one lasted three minutes; but if an infant was much distressed and could not be comforted the episode was curtailed.
Episode 5 began with mother's return, after which the stranger departed. On entering, mother was to pause in the doorway in order to see what her infant's spontaneous response to her return would be. Thenceforward she was free to do whatever suited -- to comfort him if required and to settle him afresh in play with the toys. Once he was settled she was to leave the room again, pausing briefly as she went to say 'bye-bye'.
During episode 6, in consequence, the infant was left all alone. Unless curtailed because of distress, this episode lasted the usual three minutes.
Thereafter first stranger and then mother returned to make episodes 7 and 8.
Throughout the series of episodes the behaviour of infant, mother, and stranger was recorded by observers from behind a one-way vision window. From the narrative record two measures of behaviour could be obtained for each infant: (a) the frequency with which different sorts of behaviour were shown during each episode, the frequency in every case being measured by scoring 1 for each period of fifteen seconds during which that behaviour was seen (thus, for a three-minute episode a score could range from zero to 12); (b) the intensity of certain kinds of behaviour shown during each episode; in making ratings of intensity it was often necessary to take account of how mother or stranger was behaving to the infant.
The finding to which attention is specially drawn in this chapter is that the behaviour of these fifty-six one-year-old infants during the episodes when mother was absent (nos. 4 and 6) was in every case much changed from what it had been during the earlier episode (no. 2) when mother was sitting quietly in the room with them. All the infants showed behaviour of a kind that everyone would describe as anxious or distressed, and as being due to his missing his mother.
During episode 2 while his mother was present, the typical
-41-
picture of an infant was one of active interest in the scene. As a rule he moved around freely and played with the toys, giving only an occasional glance towards his mother; a small minority (seven infants), however, were inactive and tended to stay where they had been put.
43
During this episode crying was conspicuous by its absence, though an occasional child whimpered for a few moments to begin with.
During episode 3, in which the stranger joined mother and infant, the behaviour of most of the children changed substantially. Staring at the stranger was almost universal; many infants moved rather closer to mother; and exploration and play diminished, on average to about half what they had been. Some infants showed a tendency to cry or grizzle; but in only five cases was crying of any intensity. As a rule the stranger was treated with interest and, fairly soon, with cautious friendliness.
During episode 4 mother had departed and infant found himself alone with stranger. Half the infants showed a strong tendency to seek mother, starting usually as soon as they realized she had gone. Eleven followed mother to the door or struggled to do so; the others either looked at the door frequently or for long periods, or else searched for mother in the chair on which she had been sitting. There was also much crying and other signs of distress. For the group as a whole there was four times as much crying during mother's absence as there had been during episode 3. A dozen infants cried practically the whole time and another thirteen for a part of it. In all, thirty-nine infants either cried or searched, or did both (thirteen cases). This leaves seventeen, a fairly substantial minority, who did neither.
The behaviour of the infants during episode 5, after mother had returned, is referred to again in Chapter 21. Suffice it to say that half of them actively approached mother and showed a clear desire to be close to her, while another six either signalled or approached in a less purposeful way. Thirteen of the more active ones, having achieved close physical contact with mother, maintained it both by clinging to her and by resisting her attempts to put them down. All who had been crying stopped doing so although some infants who had been acutely distressed were not comforted quickly.
During episode 6, after mother had departed again, this time leaving her infant all alone, searching and crying were seen in more infants and were also more intense than they had been during episode 4. On this occasion forty-four infants searched for mother, of whom thirty-one followed her to the door. Of the
-42-
thirty-one, fourteen banged on the door or tried to open it: they either reached vainly for the knob or tried to insert their fingers into the crack. Among the dozen infants who did not search for mother, there were a few who had searched for her during the first separation episode but who on the second occasion showed only distress. During episode 6 there was also a great deal of crying. Forty infants cried more or less strongly; these included all those who had been distressed during the earlier episode as well as many others. Some rocked, or kicked their heels on the floor, or moved at random 'like a little trapped animal'. Only two infants neither searched nor cried; thirty did both.
Episode 6 was ended by stranger's return and the start of episode 7. After three minutes, during which infant was with stranger, mother returned and episode 8 began.
During episode 8, the tendencies to approach mother, to cling to her, and to resist being put down were much stronger and were seen in more of the infants than had been the case during the previous reunion episode. This time thirty-five of the fifty-six infants actively approached
44
mother and showed an evident desire for physical contact; a further nine either signalled their desire for contact or else approached mother in a less purposeful way. Two other infants, though they did not approach mother, engaged in lively interchange with her across a distance. Most striking was the large number of infants (forty-two) who both actively clung to mother and resisted being put down by her; another three, though they did not cling, resisted being put down.
A minority of infants included in the above figures showed signs of ambivalence towards mother of greater or less degree. Thus a few were seen to ignore mother briefly before approaching her, and others alternated between approaching and turning away. There were a few others who were so ambivalent that they mingled active attempts to seek and maintain contact with mother with attempts to get away from her.
A further small minority of infants (seven) behaved quite differently: they neither approached mother nor showed any desire to do so. Instead, they persistently ignored her and refused to respond when she invited them to come. Some even avoided looking at her.
To return to episodes 4 and 6: when behaviour shown during these two episodes, when mother was absent, is examined it is found that an infant was extremely likely either to search or to
-43-
cry or to do both together. The table below shows the number of children who responded in one of these three ways for each episode.
Behaviour
Cry only Search only Search and cry Total (N = 56)
Episode
4 6
12 10 14 14 13 30 39 54
When the behaviour of any one infant in episode 6 is compared with what he showed in episode 4 the following sequences are found:
-- those who cried only in episode 4 were likely to do the same in episode 6
-- those who searched only in episode 4 were likely to search and cry in episode 6
-- those who searched and cried in episode 4 were likely to do the same in episode 6, though a few cried only.
The individual differences in the responses of these children are of great interest and (as is discussed in Chapter 21) are correlated with the different patterns of mother-child interaction observed in the preceding year. Here, however, our concern is with the features that the infants' responses had in common. On each occasion when mother left the room, first leaving infant with stranger and then leaving him all alone, the behaviour of every infant changed. Play and exploratory behaviour either slowed down or ceased altogether. On the second
45
occasion especially, all but two of the infants showed marked dislike of the situation and expressed it by seeking mother, by crying unhappily, or by doing both together. The extent of distress and anxiety during mother's absence was considerable, even though the room and the toys remained exactly as they had been.
While to describe the broad features of behaviour for the sample as a whole enables generalizations to be framed with confidence, it tends also to be a little impersonal. To illustrate something of what the series of episodes meant to one small boy
-44-
and his mother, a description is given of one case, selected to be as representative as any one case can be: 1
1. Mother, Baby, Observer. Brian had one arm hooked over his mother's shoulder as they
came into the room; he was holding on to her, grasping a fold of her blouse. He looked
around soberly, but with interest, at the toys and at the observer.
2. Mother, Baby. After being put down, Brian immediately crept towards the toys and
began to explore them. He was very active, picking toys up, then dropping them or moving them about, with vigorous movements. He crept around quite a bit, mostly on his mother's side of the room. Although his attention was fixed on the playthings, he glanced up at his mother six times, and smiled at her twice. She glanced at him covertly, from time to time, but their glances did not seem to meet. Once he threw a toy with a clatter at her feet; she moved it back towards him. Otherwise there was no interaction between them. Towards the end of the three minutes he blew into a long cardboard tube, vocalizing as though pretending it were a horn, and then he looked up at his mother with a smile, seeming to expect her to acknowledge his accomplishment.
3. Stranger, Mother, Baby. He turned to look at the stranger when she entered, with a pleasant expression on his face. He played with the tube again, vocalized, smiled, and turned to glance at his mother. He continued to play, glancing at the stranger twice. When the stranger and his mother began to converse, he continued to explore actively at the end of the room, and looked up only once -- at the stranger. Towards the end of this minute of conversation he crept over to his mother, pulled himself up, and stood briefly, holding on to her knee with one hand, and clutching her blouse with the other. Then he turned back to play. When the stranger began her approach by leaning forward to offer him a toy he smiled, crept towards her, and reached for it. He put the toy in his mouth. She offered him the tube and he blew into it again. He looked back and forth from the toys to the stranger and did not look at his mother at all.
4. Stranger, Baby. He did not notice his mother leave. He continued to watch the stranger and the toys she was manipulating. Suddenly, he crept to his mother's chair, pulled
____________________
1 This description is taken from Ainsworth & Wittig ( 1969).
-45-
himself up into a standing position, and looked at the stranger. She tried to distract him with a pull-toy. He approached the toy, and began to roll it back and forth; but he glanced again at his mother's empty chair. He was less active than he had been when alone with his mother, and after two minutes his activity ceased. He sat chewing the string of the pull-toy, and glancing from the stranger to his mother's chair. He made an unhappy noise, then a cry-face, and then he cried. The stranger tried to distract him by offering him a block; he took it but
46
then threw it away with a petulant gesture. He gave several more little protesting cries, but he did not cry hard.
5. Mother, Baby. When his mother opened the door and paused in the doorway, Brian looked at her immediately and vocalized loudly, with a quality that could have been either a laugh or a cry; then he crept to her quickly, and pulled himself up, with her help, to hold on to her knees. Then she picked him up, and he immediately put his arms around her neck, his face against her shoulder, and hugged her hard. He then gave her another big hug before she put him down. He resisted being put down; he tried to cling to her and protested loudly. Once on the floor, he threw himself down, hid his face in the rug, and cried angrily. His mother knelt beside him and tried to interest him in the toys again. He stopped crying and watched. After a moment she disengaged herself and got up to sit on her chair. He immediately threw himself down and cried again. She helped him to stand, and cuddled him. For a moment he reciprocated in the cuddle, but then he threw himself down on the floor again, crying. She again picked him up, and tried, to direct his attention to a squeaky ball. He looked at it, still holding on to his mother, with one arm hooked over her shoulder. He began to play, but quickly turned back to his mother with a brief cry, and clung to her. This alternation of play and clinging continued. After four and a half minutes, his mother, apparently not wishing to delay us, picked a moment when he was interested in a ball, and moved to the door.
6. Baby Alone. As she said 'bye-bye' and waved, Brian looked up with a little smile, but he shifted into a cry before she had quite closed the door. He sat crying, rocking himself back and forth. He cried hard, but occasionally lulled a little and looked around. After a minute and a half the episode was curtailed and stranger instructed to enter.
-46-
7. Stranger, Baby. Brian lulled slightly when he saw the stranger enter, but he continued to cry. She first tried to distract him, then offered her arms to him. Brian responded by raising his arms; she picked him up, and he stopped crying immediately. She held him in her arms, and showed him the pictures tacked up around the edges of the mirrorwindow. He looked with apparent interest; he held on to her tightly, grasping a fold of her clothing. Occasionally he gave a little sob, but for the most part he did not cry. But when she put him down, he screamed. She picked him up again, and he lulled.
8. Mother, Baby. At the moment that his mother returned Brian was crying listlessly. He did not notice his mother. The stranger half-turned and pointed her out. Brian looked towards her, still crying, and then turned away. But he soon 'did a double take'. He looked back and vocalized a little protest. His mother offered her arms to him. He reached towards her, smiling, and leaned way out of the stranger's arms and his mother took him. He threw his arms around her neck, hugging her hard, and wiggling with excitement. Then the stranger tried to attract his attention. Brian did not notice her advance until she touched him; he immediately clung to his mother and buried his face in her shoulder. His mother continued to hold him, and he cuddled and clung to her, as the episode ended.
Since Ainsworth first reported her findings the results of several other studies have come to hand. In three of them ( Maccoby & Feldman 1972; Marvin 1972) the series of experimental situations used was planned to be as similar to Ainsworth's as possible, but in each the children were older. In a further two studies ( Cox & Campbell 1968; Leeet al. , in preparation) the situations differ from those used by Ainsworth, but in each there is opportunity to study children in an experimental setting first with mother present and later with mother absent. Details are given in the table overleaf.
47
Because the last two studies listed give data comparing behaviour at one year with that to be seen at two years and onwards it is convenient to present findings from one of them first.
Preliminary findings reported by Lee and his colleagues show that, while attachment behaviour continues to be extremely active at the time of the second birthday, the behavioural system
-47-
Authors
Maccoby & Feldman Maccoby & Feldman Marvin
Cox & Campbell
Leeet al.
Ages at children studied
2, 2 1/2, 3 years 2 1/2 years
2, 3, 4 years
14 months 24-37 months 1, 2, 3 years
which
Samples
White American 30-60, longitudinal
From Israeli kibbutzim 20, cross-sectional
White American 3 * 16, cross-sectional
White Canadian
2 * 20, cross-sectional
Middle-class English 27, longitudinal
governing it has altered in many respects since the first birthday. A comparison of the behaviour of the same children placed in the same situation at one year and two years of age shows that, at two years of age, children are likely:
-- to maintain greater proximity to mother -- a finding already reported from observations made out of doors by Anderson ( 1972a)
-- to be more hesitant in approaching a stranger.
On the other hand, merely to be close to mother and to be able to see her seem sufficient to give a child of two years a sense of security, whereas a one-year-old is likely to insist on physical contact. Two-year-olds, moreover, protest less than do oneyear-olds during a brief period when mother leaves them alone. Lee concludes that, compared with one-year-olds, two-yearolds have available more sophisticated cognitive strategies for maintaining contact with mother. They make much more use of looking and verbal communication, and are probably also using mental imagery in ways hardly possible for a one-yearold. As a result their attachment behaviour is better organized and their proximity-keeping is more proficient than it was when they were a year younger.
During the third year of life changes in the behaviour seen in the experimental situations are probably also in large part a result of the developments occurring in a child's cognitive competence. In their longitudinal study of children between
48
-48-
their second and third birthdays, Maccoby & Feldman ( 1972) note the much greater ability of three-year-olds to communicate with mother over a distance and also their increased ability to understand, when mother leaves the room, that she will soon be returning. As a result, when the reaction of three-year-olds to mother's brief absence is compared with that of two-year- olds, both crying and going to the closed door show a marked decrease. In addition, three- year-olds who have been left alone recover their equanimity when they are rejoined even by a stranger, whereas two-year-olds remain as upset when the stranger returns as they were when left entirely alone.
The responses of the children observed by Maccoby & Feldman when tested at the intermediate age of two and a half years in the same series of situations were roughly intermediate between the responses seen at two years and those seen at three years. Interestingly enough, the behaviour shown in the same situations by kibbutzim children at the age of two and a half differed very little from that of the American children at that age. Similarities between the groups were found in regard both to the means for the groups and to the range of individual variation within them. These findings are in keeping with other observations that suggest that the development of attachment behaviour in children brought up in kibbutzim is in most cases very similar to that of children brought up in traditional families (see Volume I, Chapter 15).
Although attachment behaviour develops in important ways during the second and third years of life, a child's behaviour in these experimental situations when mother is absent continues to be very different from what it is when she is present. For example, Maccoby & Feldman found that the manipulative play of two-year-olds decreased by about one-quarter when they were left with a stranger and by about half when they were left alone. Conversely, the proportion of children who cried increased enormously, from 5 per cent when mother was present to 30 per cent when the children were left with a stranger and to 53 per cent when they were left alone. Changes in the behaviour of three-year-olds when mother was absent were less striking than in two-year-olds but in the same direction. Their manipulative play decreased by one-sixth when they were left with a stranger and by one-third when they were left alone. The proportion of children who cried rose from zero to 5 per cent and 20 per cent respectively in the two situations in which mother was absent.
-49-
In addition to crying after mother had departed, there were many children at each age-level who showed a desire to follow her. Of the two-year-olds, 30 per cent not only went to the door but made efforts to open it; and a further 21 per cent stood near the door or leaned against it. Of the three-year-olds, 34 per cent attempted to open the door, and nearly half of them banged on it vigorously. At each age-level, again, a substantial minority of children expressed anger at mother's absence: 19 per cent at two years, 31 per cent at two and a half years, and 14 per cent at three years.
In noting the increased activity observed when children were left alone, especially evident at the ages of two and two and a half, Maccoby & Feldman write:
This increased activity frequently took the form of anxious searching or agitated movement. There was occasionally a quite opposite kind of reaction to the stress of being alone: a kind of
49
frozen immobility. . . . some children stood very still. This might occur near the door, when the child appeared to be waiting for his mother's return, or it might occur elsewhere in the room. In a few instances the child played with the toys, but each movement occurred at a markedly reduced speed, much as though the action had been rendered on a slow-motion film. Also, it occasionally happened that a child who was upset over separation would alternate between an unfocused running activity and immobility.
As regards the evaluation of these observations it is perhaps necessary to remind readers that on each occasion when mother departed she was absent for no more than three minutes, and for an even shorter time if the child was distressed, and that on the first of the two occasions the child was left with a friendly female stranger whom he had first met in the presence of mother. Furthermore, the toys he had been playing with were still there.
The findings of a cross-sectional study of samples of eight boys and eight girls at each of three age-levels by Marvin ( 1972) are in broad agreement with those of Maccoby & Feldman and here observations are extended to the fourth birthday. In Marvin's study the behaviour of boys and girls tended to differ. The two-year-old boys were as much upset as were Ainsworth's one-year-olds. Three-year-old boys were less upset than twoyear-olds; and four- year-old boys were comparatively little
-50-
affected by any of the situations. By contrast, the two- and three-year-old girls were appreciably less affected by events than were one-year-olds, whereas the four-year-old girls were much more upset, especially by being left alone. An explanation suggested by Marvin of this last result, which is unexpected, is that a four-year-old girl may be especially disturbed by mother's apparently arbitrary behaviour in the test situation and by her unwillingness to cooperate when the child asks her not to leave her alone. Although in broad outline the findings of these different studies are consistent, there are many differences of detail. For example, neither Ainsworth with her one-year-olds nor Maccoby & Feldman with their two- and three-year-olds found sex differences of any magnitude; whereas Lee and his colleagues with their one- and two-year-olds and also Marvin with his twos, threes, and fours were struck by the differences between boys and girls. This and other discrepancies in the results reported in different studies are not easy to interpret. It seems not unlikely that relatively small differences in the arrangements for the testing, for example, in the behaviour of the stranger, can affect considerably the intensity, though not the form, of any behaviour exhibited. From these and other miniature separation experiments certain conclusions can be drawn:
a. In a benign but slightly strange situation, young children aged between eleven and thirty- six months, and brought up in families, are quick to notice mother's absence and commonly show some measure of concern, varying considerably but amounting very often to obvious, and in some cases to intense, anxiety and distress. Play activity decreases abruptly and may cease. Efforts to reach mother are common.
b. A child of two years is likely to be almost as upset in these situations as a child of one, and at neither age is he likely to make a quick recovery when rejoined either by mother or by a stranger.
c. A child of three is less likely to be upset in these situations and is more able to understand that mother will soon return. On being rejoined by mother or a stranger he is relatively quick to recover.
d. A child of four may either be little affected by the situations or else be much distressed by mother's apparently arbitrary behaviour.
50
-51-
e. As children get older they are able to use vision and verbal communication as means for keeping in contact with mother; should they become upset when mother leaves the room older children will make more determined attempts to open the door in order to find her.
f. Up to 30 per cent of children are made angry by mother's leaving them alone in these circumstances.
g. In some studies and at some ages no differences are observed in the behaviour of boys and girls. In so far as any differences are observed, boys tend to explore more in mother's presence and to be more vigorous in their attempts to reach her when she has gone; girls tend to keep closer to mother and also to make friends more readily with the stranger.
A further finding from these miniature separation experiments, and one that links with the findings of Shirley ( 1942) and Heathers ( 1954) (see pp. 35 and 37 above), is only very recently reported. This is that when a child of about one year is tested in Ainsworth's series of episodes for a second time, a few weeks after the first testing, he is more upset and anxious than he was on the first occasion. When mother is present he keeps closer to her and clings more tightly. When she is absent he cries more ( Ainsworth, personal communication). These findings emerge from a test--retest study of twenty-four babies tested first at fifty weeks of age and a second time two weeks later. On the assumption that increased sensitivity is not due simply to maturation, which is unlikely, these findings provide the first experimental evidence that at one year of age a separation lasting only a few minutes, in what would ordinarily be regarded as a bland situation, is apt to leave a child more sensitive than he was before to a repetition of the experience.
Ontogeny of responses to separation The First Year
Since the responses to separation that are so unmistakable in infants of twelve months and older are not present at birth, it is clear that they must develop at some time during the first year of life. Unfortunately, studies designed to throw light on this development are few, and are confined to infants admitted to hospital.
