- Ethics Home Page
- About the Center
- Focus Areas
- Contact Us
- Site Index
Attachment 101 for Attorneys: Implications for Infant Placement Decisions
Recently, the media has highlighted several controversial child custody battles involving very young children. In each case the primary question considered by the courts was which of the caregivers had the right to custody of the child. We consider here the more critical question: What is in the best interest of the child? We will examine this question in relation to the case of Michael H.1
Michael was conceived in July 1990 by a young engaged couple, Mark and Stephanie. Shortly after learning of the conception, the couple agreed to give the child up for adoption. During the first months of pregnancy, Mark and Stephanie cared for the unborn child together. However, as the months advanced, the stress of pregnancy took a toll on their relationship, and the couple broke up. Shortly after the disintegration of the relationship, Mark attempted suicide and was hospitalized. During this time, Stephanie continued making preparations for the infant's adoption. Also during this time, Mark resolved to make substantial changes in his life. He decided he did not want to give the infant up for adoption, and he began seeking legal counsel to help him prevent it. Michael was born in February, 1991 and placed in the care of John and Margaret Stedmond. In March of that same year, Mark learned of his son's birth, and he began making arrangements to obtain custody.
What will happen to Michael as the custody battle ensues, and how will his intimate relationships be affected? How should we, as members of society and investors in future generations, support Michael and help cultivate his attachment relationships? The purpose of this paper is to provide background for an informed consideration of these questions. We begin with the concept of attachment.
II. Secure and Insecure Attachment
A. What is Attachment?
Intimate relationships with others -- like the one Michael formed with the Stedmonds -- are the context in which we discover who we are, learn how others feel about life's important issues, and find out how to bridge differences. The emotional security and warmth derived from an initial close relationship with a loving parent provides us with a "home base" from which we can venture to take the risks that are inevitably part of a life of joy and accomplishment. In short, close, psychologically intimate relationships between babies and their caregivers are central to human life. The theory of attachment is about these relationships; how they are formed, what happens during the first intimate relationship with the nurturing parent, and what the consequences are for later development. We present this theory along with a summary of the research and clinical evidence that supports it. Using this theory as the basis of discussion, we develop a case asserting a societal duty to protect these relationships of attachment as part of the general responsibility to protect the very young.
The longing that we all feel for intimate closeness with another human being is moderated, often greatly, by the competing need for autonomy and separateness. Attachment theorists regard this tension between the needs for closeness and autonomy as the central dilemma of human life. The theory, as expressed in the writings of John Bowlby2, develops the interplay between attachment and independence. A satisfying close relationship with a nurturing caregiver -- a secure attachment -- provides the base of operations from which a toddler can venture forth into independent activity. Throughout the life span, one continues to feel the rival tugs of closeness and autonomy. If we begin life with an experience of successful closeness, we are not only better able to create closeness in our relationships with friends and partners, but we are also more equipped to take the risks involved in having a sphere of separateness for ourselves.
John Bowlby, the main theorist in the attachment literature, identifies the close attachment relationship between responsive caregiver and dependent infant as an adaptation needed for the infant's survival. The emotionally charged connection between caregiver and child ensures the two will remain near each other physically, at least while the child is an infant. At the same time, the security of this very closeness creates the courage the child needs to venture forth away from the close involvement she maintained earlier with the caregiver. This fascinating paradox -- between creating a close intimate connection and moving away from dependence on that connection -- is the heart of attachment theory.
The attachment we experience in our first close relationship, usually the mother-infant relationship, forms the foundation for much that happens subsequently. We form a mental representation of our experience of being loved and cared for in an intimate context. Before this image is formed, we must be physically near the caregiver to retain the feeling of security required to support play and independence. After the mental image is formed, we create a portable reminder of the closeness, love, and security we enjoyed. Thus, we can separate from the caregiver more often and for longer periods. As adults, what remains with us from our earliest attachment is a system of beliefs, images, and emotions about ourselves in loving relationships. In an ideal scenario, these cognitions denote ourselves as loved and loving. Having experienced an ideal love at least once, we approach new human connections with faith in love's possibility.
However, early relationships between caregivers and babies are not always ideal. The theory of attachment addresses how a relationship of poor quality -- an insecure attachment -- emerges from interaction that is not responsive to the baby's needs. The theory explores consequences of such an insecure attachment for the beginnings of self-awareness and the ability to be independent. In addition, recent work takes the process forward into adulthood and the formation of romantic relationships. To date, most of the material regarding how to "fix" either a poor ongoing mother-infant relationship, or the later consequences of one that occurred in the past, appears in clinical or intervention literature. The theory contains a rich set of ideas about the impact of an insecure attachment history, and implies that the impact would be hard to counteract. Yet, clinical evidence suggests that people can and do overcome the negative effects of insecure attachment under the right circumstances.
B. How Does Attachment Need Differ from Related Social Needs?
Attachment, as a motivational system, is often confused with two other human needs: dependency and sociability. But these two motivational systems are distinct from attachment in important ways that affect how attachment is understood. Dependency is the need to obtain the assistance of other people in order to meet other needs. These other needs may be social, such as attention or companionship during an activity, or physical, such as food or transportation. The social partners are needed because they can help meet other needs, not as an end in themselves. Dependency arises out of our experience or concern with not being able to satisfy our own needs. In contrast, attachment arises out of an inborn need to be close, proximate, and intimately involved with one who is responsive to what we are communicating. Attachment is an end in itself whereas dependency is a state that leads to other ends. A person may be attached to another on whom they depend, but, once an attachment relationship develops and is mentally represented, a person can also be attached to someone on whom they do not depend.3
Another difference between attachment and dependency concerns the cognitive elements. The experience of a secure attachment results in a mental representation of that experience which becomes a stable element of self representation. Dependency experiences may or may not have mental representations; it is not a defining feature.
A second need that is often confused with attachment is a generalized need for companionship and social stimulation which we can call a need for sociability.4 Babies, like other people, are social creatures and often laugh, smile and express interest in the people around them. We might say they like particular people and have real social relationships with these people. However, a baby is not attached, in the technical sense, to everybody they like. Attachment involves a high degree of intimacy with a particular person and the facilitation through that intimate relationship of exploration and attendant other development. However, sociability is a need for social contact per se and can be satisfied by any number of reasonably warm and attentive people. While babies enjoy a variety of social contacts, there is an absence of reciprocity between partners and a willingness to explore that we see with attachment.
Attachment is a psychologically powerful connection between two people who participate in an intimate relationship involving care of one partner by the other or frequent response to one partner by the other. The theory presumes that there is an evolution-based drive to achieve physical and psychological closeness in just such a caregiving relationship. This drive may be related to the dependency and sociability facilitated in the close relationship, but it is not reducible to them. Attachment need is focused on the particular relationship partner who is responding to one with whom one has intimacy.
C. The Context in Which Attachment Theory Was Developed
Bowlby's first professional experience with attachment occurred during a remarkable social experiment.5 During World War II, in the Battle of Britain, English cities were bombed daily. The English decided to evacuate their children from the targeted urban areas to the countryside where they were cared for at the estates of lords and ladies. Many health professionals moving to Britain as refugees from Hitler's Germany were available, along with British child welfare professionals, to plan and staff this program. Parents readily agreed overnight to separate from their children in order to be assured of their safety.
Professionals had provided every available resource and professional planning possible. Many of the children, especially the younger ones, became withdrawn, uninterested in play, and generally apathetic. These children often huddled together without displaying any genuine social exchange. They were what we would call "depressed." These children also had medical problems of various kinds which were unresponsive to medical care.
Bowlby was pivotally involved in finding the solution to these problems. The solution involved returning the children to their parents as quickly as possible -- and sooner than originally envisioned. During the period of the evacuation, Bowlby made several observations that were central to the theory he was to develop. First, he noticed that there seemed to be no connection between meeting these children's physical needs for food and shelter and meeting their psychological needs. This lack of correlation between attachment and physical need satisfaction was at odds with the psychoanalytic orthodoxy of Sigmund Freud. According to Freud, attachment evolves from dependency; however, Bowlby observed that this simply was not true. These children were not becoming close to and secure with their physical caregivers.
Bowlby's second observation was that these children were not forming any new attachments with caring adults. Thus, it could be deduced that what such attachments depend upon was apparently not present. The children seemed unable to care enough about the details of their caregivers' comings and goings to form a connection. In addition, the children initiated little communication, even in the form of crying. It was as if these children had lost faith in their capacity to influence anyone else.
Bowlby's last observation during the Battle of Britain evacuation regarded play. He noted with alarm that these children were not playing. This revelation planted the seed for the idea that attachment might paradoxically be connected to a child's independent behavior. Since he had been trained in psychoanalytic theory, Bowlby was not initially convinced of the importance of emotional needs not derived from biological drives. However, he had studied and admired Darwin's theory of evolution, and soon he began seeing a connection between this theory and his own observations of the evacuated children.
D. Two Major Influences on Bowlby's Theory of Attachment
Two major theories influenced Bowlby as he struggled to make sense of his observations of the evacuated English children. The first was classic psychoanalytic theory as presented by Freud which formed the core theoretical background in Bowlby's psychiatry and child development training. The other important influence was Darwin's theory of evolution, which Bowlby had encountered throughout his education and which had intrigued him greatly.
The classic psychoanalytic perspective on human nature was built on four key ideas. First and foremost was the idea that human nature revolves around innate biological drives, specifically the sex drive and the aggression drive.6 Second, Freud postulated that much of the time our biologically driven motives operate outside our awareness. Third, Freud perceived our conflicts regarding these unconscious biological drives to be the source of our development. Thus, in order to mature properly, we must learn to channel our driven behavior into realistically adaptive and socially acceptable paths. Finally, Freud asserted, based on a physical science model, that psychic energy is finite in supply. This conception, sometimes referred to as the hydraulic principle, asserts that if energy is expended in one direction, there will be that much less energy available to expend in another direction. For example, if an infant spends a great deal of energy crying to be fed, it will have less energy left to communicate with the caregiver.
In conjunction with his clinical experiences with disturbed adults, these ideas allowed Freud to design a theory depicting human development as a matter of distributing psychic energy among various biologically ordained but conflicting goals. Thus, attachment between child and caregiver was viewed as a byproduct of the child's struggle to meet its own sexual and survival needs. From this perspective, it was surprising that the evacuated children in World War II England failed to develop attachments to those who provided for their basic survival. Because of Freud's pervasive influence on society's mainstream beliefs about the parent-child relationship, people often expect that any caregiver who regularly meets a baby's needs will become an attachment partner for that baby. It is important to note that the evidence does not support this connection between a caregiver's physical ministrations to a child and the formation of an attachment relationship between that caregiver and child.7
Bowlby's observations of the children's responses to the disruptions in their parent-child relationships led him to question the idea that the attachment need derives from other needs. It seemed clear to Bowlby that the evacuated children needed to be with their mothers and were suffering because of separation. He began to integrate his observations with Darwin's writings regarding the inheritance of motives.8 From an evolutionary perspective, needs evolve when their satisfaction will be adaptive and will ensure species survival. Proximity to, communication with, and physical contiguity with a caregiver are all objectives that facilitate care. Thus, Bowlby postulated that a drive -- or evolved motive -- to obtain such proximity, communication, and closeness exists to ensure human survival. This attachment need may facilitate the meeting of needs identified by Freud -- survival and sexual expression -- but is separate from them.
E. Key Ideas of Attachment Theory
Bowlby's theory of attachment presents several cardinal ideas.9 Together, these ideas constitute a powerful general principle: human relationships drive human development. We have organized Bowlby's discussion of these key ideas into four general principles.
I. Attachment is a Biologically Based Need Separate from Other Such Needs
Attachment is a motivational system separate from others such as those for food, water, sex, and social connection. Attachment involves an emotional sense of secure closeness to the caregiver. In infancy, actual proximity, physical contact, or communication through eye contact or gesture is required to produce this feeling of closeness. Thus, when separation or loss of contact occurs, the infant experiences its need for attachment and becomes motivated to seek proximity or contact using communication as one of the tools. When the feeling of closeness is restored, in other words, when the child feels "refueled," the seeking of proximity and closeness recedes, and the child turns to other interests.
We see in the above analysis that the attachment motive and attachment behaviors are two distinct entities. The motive is a need for closeness with a particular partner with whom the child is involved in an intimate relationship. The behavior -- proximity seeking and so on -- is evident only when the need is not being met. An important implication of this fact is that we cannot tell much about the intensity or quality of a particular child's attachment by observing only its efforts to get nearer to or have contact with the mother. An evaluation of a particular mother-child attachment relationship requires us to observe the pair in disquieting situations that put the availability of the mother in question (in the child's mind) and in more comfortable situations where the availability of the mother is clear. Only then can we determine if the frequency of attachment behavior varies with the situation.
2. Attachment Need and Exploratory Behavior Have a Reciprocal Relationship
The second key idea of attachment theory also regards the variations in behavior in situations that differ with respect to mother availability. When secure in the presence of its mother, the child meanders off to explore the toys, people, or other interesting materials around. During this time, an abundance of exploratory behavior and little or no attachment behavior is displayed. However, if the child toddles off too far or plays intently for a while, concern about separation from the mother builds until the child stops exploring after a few minutes and returns to "refuel" the attachment need by being close to and communicating with the mother. After checking in with its mother, the child leaves to explore again.
This pattern of alternation between attachment behavior and exploratory play is called "the secure base." When a child is securely attached to its mother or other regular care provider, the child is able to use that person as a base of operations from which to explore. Paradoxically, we see more independent exploration of the environment in children who are closely and securely attached than we do in those who are insecure in their attachments. The comfortable use of the caregiver as a base of operations is a central idea of attachment theory.
3. Each Person Forms a Mental Representation of the Early Attachment Experience
Bowlby follows psychoanalytic theory closely by presuming that securely attached children form mental images of their ongoing relationships with their mothers. Subsequent research, based on a structured interview method more reliable than the psychoanalytic approach, has confirmed the existence of these representations.10 This mental representation of the attachment relationship is called an "Internal Working Model of Attachment." The working model is dynamic; in other words, it is a representation of the ongoing and varying experiences of being cared for and recognized by the mother. An essential component of this complex mental representation is an image of the self as someone who is and can be cared for and loved. A reciprocal part of the representation is an image of the mother caring for the child. In attachment theory, this mental model of the attachment relationship is carried forward in development and is the foundation for either our faith in love's possibility, or our doubts about that very issue.
Once we form a clear mental representation of self and caregiver in relationship, we no longer need the concrete physical reassurance the younger child needs. Instead, we can carry a mental image inside us as we explore, go to school, meet new people, and so forth. Thus, internalization of a close emotional bond actually facilitates our growing independence. The mental model is also the foundation of our capacity to form, sustain, and commit to close relationships in the future.
4. Awareness of the Self as an Autonomous Person is Facilitated by the Close Intimate Connection We Call "Attachment"
An important feature of the working model is its recognition of the self having impact on another. Part of the model, then, is a representation of the self as an effective influence on others. Attachment theory also illuminates the fact that the origin of self-image is the perception of self in relationship to others.
Combined, these four ideas summarize attachment theory. Attachment is a basic human need for a close, intimate relationship with at least one other person. At first, it is expressed concretely by the child seeking proximity to and physical contact with the caregiver. Later, after the mental working model has been formed, the child is able to tolerate separation due to its portable reminder of the relationship. Thus, secure attachment actually facilitates exploration and independence as well as the formation of an autonomous self and new relationships.
F. How Do We Observe Children's Attachments in Practice?
Mary Ainsworth, an American developmental psychologist who studied with Bowlby, developed a controlled laboratory situation that allows researchers to classify children according to the quality of their attachment to the parent or other caregiver.11 Ainsworth's procedure is called "The Strange Situation" because it is a novel experience for the child. Ainsworth modeled her design for the analysis on the patterns of comings and goings of parents and strangers that are commonplace in children's lives. The child is observed playing for twenty minutes in a small playroom through eight episodes of increasing distress. The episodes begin with only the mother as a companion and proceed through the introduction of a stranger, the departure of the mother, being alone, the return of the stranger, and the reunion with the mother.
This "Strange Situation" procedure allows us to observe how two kinds of behavior -- attachment and exploration -- increase or decrease in intensity as the situation changes. A child who is securely attached to its caregiver will explore the room and the toys it contains when the mother is present, continue exploring but with less involvement when the stranger enters, explore minimally when alone with the stranger, and perhaps not explore at all when entirely alone. Gradually, exploratory play resumes when the stranger returns, but it is only after the final reunion with the mother that play is restored to the level it was at the beginning of the procedure. Thus, even with a baby who does not cry when separated, we have a barometer of how it responds to such events. Attachment behavior reveals a reciprocal pattern: a baby performs little proximity seeking and protesting when playing with the mother present, but protests the mother's departure and greets her avidly upon reunion.
Based on observations of this reciprocal play pattern, Ainsworth established three different categories of attachment relationships.12 The first category is secure attachment. The securely attached infant actively explores while alone with the mother and is visibly upset by separation. The infant greets the mother warmly upon return and welcomes physical contact with her. The child is outgoing with strangers while the mother is present.
The second category of attachment is anxious-ambivalent insecure attachment. The anxious-ambivalent child appears anxious and is unlikely to explore while the mother is present. In addition, the child becomes extremely distressed when its mother departs. Then, when the mother returns, the child reacts very ambivalently, trying to remain near the mother, although resenting her for having left. The child is also likely to resist contact initiated by the mother. The child is quite wary of strangers, even when the mother is present.
The third category of attachment is anxious-avoidant insecure attachment. The avoidant infant appears uninterested in exploring when alone with the mother. Moreover, the child displays little distress when separated from the mother and often avoids contact with her when she returns. The child is not particularly wary of strangers, but may avoid them as it does the mother.
To illustrate these attachment categories, we will look at Sarah, Julie,
and Amy -- three hypothetical one-year-olds regularly attending day care.
The classifications made by clinicians trained in attachment theory have been robust in their ability to predict other aspects of mother-child interaction in the manner consistent with attachment theory.13 For this reason, it is becoming increasingly common for clinical evaluators to look at attachment issues when working with the under three age-group. What is distressingly uncommon, however, is for social workers and attorneys involved in the placement of children up to three-years-old to be looking at or taking account of the attachment relationships of these children.
G. How Do Secure Attachments Get Formed?
Mary Ainsworth's original classic research followed a group of twenty-six families over the first year of the children's lives.14 She had home observers visit the families four times for a number of hours, and numerous interviews were conducted as well. The Strange Situation test was performed within a week of the child's first birthday. Ainsworth observed that the mother-child classification of securely, insecurely-avoidant, or insecurely-ambivalent attachment was related to many measures of their interaction over the first year of life. What emerged was a clear finding which has been replicated many times since this first study.15 The use of a mother's caregiving style that Ainsworth and the rest of the child development community deems "responsive" is associated with the infant's secure attachment at one year.
Responsive mothers are mothers who give physical care, emotional communication, and affection to children in clear relation to their signals of need.16 They are also giving their children "space" when they are playing, crawling, or otherwise doing their own thing. The critical aspect of responsive caregiving is that the mothers are aware of the child's needs and wants and are responding to that information. Less responsive mothers may do a lot of nursing, holding, or playing, but only in relation to their own needs rather than the child's needs. Another kind of unresponsive mother is simply disengaged from the child. The involved but nonresponsive mother is associated more often with the avoidant style of attachment in which the baby plays steadily through episodes of stranger entry, mother departure, being left alone, etc. The uninvolved mothering style is associated with the ambivalent attachment pattern in which the baby first makes an effort to connect and then withdraws, and may alternate these responses through several cycles.17
H. What Does Being Securely Attached Do for a Child?
The experience of being securely attached to at least one parent or important other caregiver in the first three years of life yields enormous dividends for human development. Indeed, this early relationship can be said to drive development. This is a strong statement and requires some justification in relation to the evidence. Fortunately, the literature relating early attachment to other aspects of development is extensive. Let us take a look at the ways in which the attachment experience affects what happens to us throughout the different stages of life.
1. The First Three Years
The first impact of secure attachment is paradoxical on the surface. Children who are securely attached at one year are better able to explore alone, separate when necessary, and operate independently from the caregiver.18 Independence is the legacy of the close interdependence that is secure attachment. Theoretically, the securely attached child has formed the mental image of the self as a cared about person in conjunction with the mother or other partner as a caring person. This mental image maintains a form of emotion regulating function that enables the child to tolerate being separated from the caregiver, both physically and mentally, without anxiety increasing enough to disrupt play. The evidence now available confirms that securely attached toddlers are more independent than insecurely attached children, but this evidence does not reveal the direction of cause and effect, and it does not yet confirm that emotion regulation is the process that facilitates independence for securely attached children.
Another outcome of a positive experience with attachment is the beginning of self-awareness.19 After the child experiences this base of operations pattern for awhile, there begins to form a mental representation of the child and its mother as two separate but closely related people. Sometimes the mother is responding in this image and sometimes not, but she responds often enough for the child to have an internal representation of a "good self" and a "good mother." Once these mental images are formed, there is less need for the frequent physical proximity between mother and child. The child now has an internalized picture of responsiveness that can be carried around for security. Thus, we see that a solid sense of self, as well as a picture of the other as a separate person, are the immediate but permanent products of the attachment relationship in the infant and toddler period of development.
After the mental representations of the cared for self and the caring parent develop, the next step is the realization that the self and the caregiver are two separate people. The representation of self begins with the child's sense that it is making things happen, that it is having influence. This is what psychologists call a sense of agency.20 The child with a clear sense of agency then proceeds, as cognitive skills increase, to understand that this self who affects influence belongs to various categories -- "girl," "three-years-old," "Smith family" -- and has various traits -- "good," "big girl," "helper." Securely attached children show better defined self-awareness than those that are insecurely attached.21 Thus, one function of attachment is the provision of the base from which we define ourselves.
An additional effect of a secure attachment on development is in the
intellectual domain. Children with a history of secure attachment develop
more superior sensorimotor skills than those with a history of insecure
attachment. Thus, secure children are better able to master the environment
using the senses, then perform related actions than insecure toddlers.22
Let us return to our example children to obtain a better understanding
of the effects of attachment for an infant or toddler.
2. The Preschool Years, Ages Three to Five
As development continues in the preschool period where basic concepts like color, shape, and number are mastered, the advantage of having been securely attached continues. Bowlby's theory provides us with a straightforward understanding of how this link between emotional development and cognitive development forms. The toddler in a secure attachment relationship spends less time checking out where mom is and more time exploring the environment. That same child, the securely attached toddler, is able to obtain feedback on her discoveries -- colored leaves, light switches, the cat's tongue -- in her brief "refueling" interactions. It is this combination of intense exploration with social exchange that fuels cognitive advancement.
In addition, language development occurs earlier and more effectively in the securely attached preschooler.23 A toddler who develops communication skills in the context of a warm, supportive, and intimate relationship will discern the message that there is a connection to be maintained. Then, as the child discovers the power of words to give voice to one's thoughts and perceptions, it will naturally direct those words to the intimate partner. The responsive caregiver "resonates" with the child's messages even though they are difficult for others to decipher at first. The securely attached child is able to separate, but then reconnect, through verbal communication. Simply put, we are more likely to learn to talk when we have faith that somebody wants to hear what we have to say. Attachment probably has its impact on language development largely through this motivating effect. A secondary route of influence is through the securely attached child's greater exposure to important verbal messages which have been directed to it over the course of its relationship. Much evidence in the language development literature points to the importance of adult communications that are child-centered and near but just above the complexity of the child's own developing language. Particularly critical are "conversations" on topics the child initiates, even if nonverbally.24 Usually, the securely attached child has more language-rich experiences than the insecurely attached child does.
Another area of development on which attachment has demonstrable influence involves a child's social relations. Extensive literature documents that securely attached toddlers are better able to play with their peers nondestructively, develop into more socially skilled preschoolers, and go on to be successful negotiators in the more complex play of the early grade school child.
As our example children move from the toddler stage into preschool,
we witness these differences in cognitive and social development.25
3. Middle Childhood
Children's symbolic, artistic, and story-based mental representations of their early attachment experiences continue at six years of age to reflect the earlier attachment classifications, exhibiting some continuity across the years of infancy, toddlerhood, and early childhood.26 The earlier category, usually assigned between one and two years of age, continues to predict a child's well-being on a number of counts. School adjustment and early reading skill are both related to attachment at approximately one year of age.27 Peer relations, as well as relations with teachers, in the early school years are better for children with a history of secure attachment than for those with an insecure history.28
As the middle years continue, early attachment quality continues to predict important features of school and peer adjustment. Those with secure histories stand out from the others with their stronger academic aspirations, persistence, ability to elicit positive responses from teachers and avoid problem behavior, and actual school achievement. Securely attached toddlers also proceed, in middle childhood, to be more empathic, better at complex negotiation, less aggressive and withdrawn, and basically more socially competent with peers in general than those who are insecurely attached. Such securely attached children are also more independent of adults in appropriate fashions.
Returning to our example children, we can better comprehend how the
differentiations among the girls are more pronounced as they continue
to grow older.
By adolescence, the mental models people possess of their parent-child attachment experience can be studied through structured interview or questionnaire procedures. Retaining one or another kind of insecure working model of attachment during adolescence has been associated with several indications of nonoptimal development. Compared to adolescents with secure mental models of attachment, adolescents with insecure models have less adequate social adjustment as seen by their peers29 and by researchers30, less internalization of behavioral standards, less skill with regulating their own emotions, and less resiliency when they must cope with stress.31 In sum, an adolescent individual's early experience of attachment, as represented in the mental model, continues to predict that individual's success in social relationships and adapting to life's challenges.
To better portray the influence of the early attachment experience on
the social relationships of adolescence, we again return to the lives
of our example children.
5. Adult Attachment
The continuing influence of an individual's early experience with attachment in adulthood has been studied with structured interviews and questionnaires.32 These adult measures focus entirely on the mental model. It is especially important for those interested in policy to understand the distinction between a mental model of a psychological experience -- here, it is one of being cared for in the context of intimacy -- and direct memory of that experience. The reader might be appropriately skeptical about the accuracy or mental availability of any memories of early care. Thus, the research with adults utilizes only the mental representation of the experience of early care constructed by the person.
The measurement techniques used with adults allow us to classify the adults into attachment categories that are similar to the ones used to classify mother-baby pairs in the laboratory. West and Sheldon-Keller33 have identified these categories and the characteristic styles of relating used by adults in each classification. Adults whose working models are analogous to the insecure-avoidant category in Ainsworth's paradigm34 demonstrate compulsive self-sufficiency, avoiding dependence on romantic partners, friends, or family. Those with a model similar to Ainsworth's anxious-ambivalent pattern may compulsively care for their partners, friends, and families.
The patterns of adults' internal working models of attachment have been correlated with the strength of these adults' romantic relationships. Security of the mental model predicts higher relationship satisfaction and the use of one's romantic partner as a secure base in a laboratory interaction situation. The connection between security of the mental model of attachment and relationship satisfaction35 is present to a much greater extent for women than for men36, suggesting continuity from one intimate relationship to another for women, but not men. For men, relationship satisfaction may pertain more to current life events than with the their thoughts (mental models) about attachment; whereas for women, thoughts about the relationship conform to thoughts regarding earlier important relationships.
Beyond romance, the security of mothers' internal working models of attachment has been used to predict the secure or insecure category of the infant attachment formed by the mothers with their own infants.37 Research has found that parents with insecure models recall their own parents less well than other parents38, which may indicate a lack of any coherent mental representation of good parenting. As a result, mothers with insecure attachment representations are much less likely to be sensitive to their babies' cues than mothers with secure representations.39 In fact, research findings implicate insecure attachment representations -- due presumably to maltreatment in infancy -- in physical abuse of infants and young children by their parents.40
Romantic relationships and the parent-child relationship are the most studied aspects of adult life that attachment significantly impacts. We also note, however, that secure mental models have been associated with positive self-esteem41, self-disclosure in one-on-one situations42, more cooperative workplace interactions43, and greater job satisfaction44. The adult literature demonstrates a continuing influence of the mental models we developed of our early attachment relationship.
To conclude our understanding of the life-long attachment experience,
let us see how the lives and relationships of our example children culminated
I. Summary of the Importance of Early Attachment
We do not yet have definitive evidence that securely attached and insecurely attached children do, in fact, grow up to become adults with corresponding mental representations; however, there is indirect evidence that they do.45 It is becoming more and more clear that early attachment experiences are the primary learning ground upon which one learns how to relate to other people. In the case of love, we learn in our early attachments to have faith in love's possibility for ourselves. This faith includes the belief that we are lovable and the essential parallel belief that we can love. These beliefs encourage us to keep faith in important intimate partnerships when the going gets tough. Concurrently, as we learn to love intimately, we develop more general skills of attending to other's thoughts and feelings which enables us to establish better social relationships.
Attachment is given enormous emphasis by child development professionals because of its powerful impact on all aspects of human development. Attachment in infancy gives the individual a base of operations from which to venture forth to learn about the world, connect to other people in it, and acquire a firm sense of one's self and one's place in that world.
J. Can a Child Have More Than One Secure and Important Attachment?
A frequent concern in discussions of a particular baby's placement is whether or not the baby will be confused by too many caregivers and whether or not her attachment to a parent will be disrupted by close involvement with other caregivers such as day care providers, a father with joint custody, or a foster mother. The evidence reveals that babies clearly can and do form more than one attachment relationship.46 They can, for example, be attached securely to mother, father, and regular caregiver. We do not yet know the upper limit regarding how many close attachments infants can form, but the intimate and responsive nature of the caregiving relationship that produces secure attachment would suggest the number is not great.
As mentioned previously, researchers studying attachment classify their subjects into categories which include secure, insecure-avoidant, and insecure-ambivalent. When children are observed in the "Strange Situation" with their mothers and fathers separately, we detect no relationship between the category assigned with the mother and that with the father. It is possible for infants to be secure with mother and insecure with father, insecure with mother and secure with father, secure with both parents, or insecure with both parents. What we do know is that a secure attachment with at least one caregiver seems to buffer a child from the poor development we might otherwise see following insecure attachment with others. This buffering effect can be critical for the development of a child in foster care who was unable to form a secure attachment relationship with a biological parent who abused or neglected her.
The buffering can only happen, however, if the natural consequence of responsive intimate caregiving is allowed to transpire. That is, if attachments that occur between responsive foster parents and needy babies are supported by social service workers. A great deal of harm is done when social service professionals uneducated in attachment theory act in order to prevent babies from forming attachments to caregivers in the belief that such attachments must later be given up for normal development to occur. All of the evidence points to the opposite conclusions: The child needs a secure primary attachment as a base from which other relationships can develop, and the child can sustain the first attachment while forming another.
III. Attachment as an Issue for Public Policy
From this brief review of the extensive literature on the special intimate relationships of attachment that babies and toddlers develop with their caregivers, we see that these relationships are essential to the healthy development of any child. Attachment is the laboratory of human connection, the experience that prepares us for a life in which we have the opportunity to thrive. As part of our general duty to protect the young, we must support them in these early important relationships and do whatever we can to preserve the family relationships that support them. In this section, we present four key sets of ideas. First, we introduce the concept of developmental risk and argue that disruption to ongoing attachment relationships constitutes such risk. Second, we contrast the implications of two alternative policies society might create for resolving disputes about the placement and legal custody of children from zero to three years of age. One policy is based on the biological ties between the disputing parties and the child and the other is based on the attachment relationship(s) between them. Third, we propose specific guidelines for resolving disputes about placement and custody of these young children. Finally, we apply our guidelines to the case of Michael H. in order to illustrate their effectiveness in practice.
A. Disruption to Attachment as Developmental Risk
The concept of developmental risk factor is a statistical one. A child who has a risk factor is a member of a group of children for whom the percentage who will go on to develop an illness, poor mental health, inadequate school achievement, unsuccessful social relationships, etc. is higher than the percentage who will develop such problems in a group lacking the risk factor.47 The development of any one human being is not perfectly predictable from one event, even one as powerful as the loss of early attachment. We are not arguing that every single child who experiences an abrupt loss of an important attachment partner will necessarily grow up mentally ill, underachieving, or unable to relate to others. However, we are making a related point: When circumstances cause the disruption of a baby's or toddler's early attachment(s), that child is at a much greater risk for those negative developments than he or she would be without that disruption. When the policies of the state create the disruption, the state is placing its most vulnerable citizens at risk.
What is it that these children are at risk of developing? When they experience disruption of attachment, children's mental models of attachment become insecure, which means they develop difficulties with entering into new intimate relationships that they could use as secure bases for exploration, risk-taking, and perception of themselves as separate persons. Attendant to these insecure working models are all the characteristics we have discussed for children with such models -- less developed social skills, lower levels of communication skills, and less mature cognitive development. Over time, the individuals with a history of insecure attachment models become adolescents and adults who are less successful in romantic, family and other relationships than their peers with secure models.
In sum, when we deliberately remove a very young child from the only home he or she has ever known, we move that child into the group of children who mentally represent close relationships as untrustworthy. In other words, when the state places any child at risk of failing at human connection, we are effecting an irreparable harm.
B. Contrast a Placement Policy Based on Attachment Relationships With One Based on Biological Relationships
In most jurisdictions today, the biological relationship between a child up to three-years-old and a candidate for legal and physical custody is given great weight, whereas the existing intimate bond between the child and a candidate is given less weight. The American preference for basing custody decisions for the very young on biology rather than on relationships is based on three ideas about parent-child relationships. First, our society assigns to biological parents a presumptive right to form and maintain a nurturing parent-child relationship with the child. In other words, the biological parent's right to custody is recognized, whereas any right the child may have to continue any nurturing relationship he or she has developed with someone other than a biological parent is not recognized. Either we do not believe children this young actually have psychologically important relationships, or we recognize that they do, but see them as readily alterable.
Second, we believe the best interests of children will generally be served when they are living with and cared for by their biological parents. Therefore, we do not feel the need to evaluate the effects of placing a child with his or her biological parent barring the existence of clear evidence that the parent is inadequate. The cultural assumptions behind this belief include the idea that the biological tie between us and our children motivates us to care for them in accordance with their needs, and that their emerging awareness of our biological tie as they grow up will foster a stronger sense of self or identity.
The third belief underlying our current preference for biology over attachment is part of a more general belief that infancy is a time of little permanent psychological importance. We acknowledge that a change in custody might be upsetting temporarily, but we think "she'll get over it," or "he'll like his new home." The effects of disrupting an attachment are presumed to be temporary and reversible with adequate care.
If we substituted a policy based on greater respect for existing relationships rather than on biological ties, we would be doing so because of beliefs contrary to the ones outlined above. We advocate such a changed policy which we will call an "attachment-centered" policy. This new policy would be based on the idea of the baby as a person involved in relationships. The baby's right to continue in these close relationships would take precedence over the biological parent's rights to form a relationship with the child, because the child is more vulnerable and because an existing relationship is more important than a potential one.
The attachment-based policy we advocate is founded on the knowledge base we just reviewed and our belief that the continuation of early attachment relationships is in a child's best interest since these relationships provide the secure base from which healthy social, cognitive, and mental heath related development proceeds. A corollary belief is that disruption to such relationships is a serious risk factor. Such disruption destroys the child's positive mental representation of relationships and, thus, puts the child at risk for failure in future as well as present relationships.
A policy for infant/toddler placement that centers on issues of attachment will better serve the developmental needs of individual children than one based on the rights of biological parents. Such a policy will also better serve the need of society for its members to be capable of strong, robust relationships with each other in romantic contexts and family settings, as well as in the educational and workplace spheres.
C. The Judges' Dilemma
Existing law is often written in terms of competing rights of adults and, depending on the jurisdiction, either ignores the interests and rights of children -- especially those under three-years-old -- or subordinates the interests of children to those of adults. Judges who are knowledgeable about young children and their relationships are thus unable to apply this knowledge in their decisions. That is, they are unable to apply it unless they identify a new right -- the right of a child to maintain the nurturing, care-providing relationships they have developed. The cases of contested adoption, disputes over post-divorce housing for very young children, and debates about foster care policy bring this controversy between parent's rights and the child's need for nurture into focus. Something like a "right to nurture" or "right to continued family relationships" must soon be identified.
D. Proposed Guidelines for Utilizing Attachment Theory in Infant Custody Cases
The evidence we reviewed offers several implications for cases involving custody of babies and toddlers, and we have identified four general principles to guide the judicial process. Each principle leads to one or more specific recommendations for judicial practice.
1. Principle 1: Time is of the Essence
Babies' lives move swiftly, and attachment relationships are continuously forming. Disruption of intimate caregiving relationships becomes more and more problematic with passing time. Periods of time allowed for disputing parties to act must be set in relation to the baby's age. We recommend the following policies as minimally adequate to protect babies' security:
2. Principle 2: Decision Makers Must Be Educated About Aspects of Child Development Discussed in This Paper
3. Principle 3: Respect for the Personhood of the Very Young Child
Equal protection by society through its laws must be given to every baby who requires society's intervention in his or her life. A key aspect in our personhood is our ability to form and maintain relationships. It is urgent that we both respect and protect the relationships into which babies have entered. Ideally, one legal procedure would exist for all matters of child placement either after divorce, after dependency petitions by the state have been granted, or after a biological parent disputes an adoption.
4. Principle 4: Best Interest and Detriment Issues
Best interest and detriment issues must be resolved in terms of the
scientific evidence regarding the impact on the child's future development
of various alternative scenarios.
Children have the right to the intimate relationships that provide them security, and it is our social responsibility to protect this right. Thus, we must offer our collective societal encouragement to those courageous enough to offer them love and commitment in adoptive families.
Unfortunately, four years later, the custody battle for Michael H. still rages. Michael lives with the prospective adoptive parents, the Stedmonds, who are struggling to maintain custody of him. Mark has spent the past four years proving his status as a presumed parent. After two appeals, Mark has won this leg of the race. However, the issues of Mark's fitness to parent and custody placement still need to be addressed. Thus, there could be another four years of courtroom contention before Michael's custody placement is finally determined. By that time, Michael will be eight-years-old and will have intensified his primary attachment with the prospective adoptive parents. Thus, if custody is granted to Mark, Michael will have been robbed of his first attachment relationship.
We propose that the father's right to custody and the infant's right to form secure attachments could both have been honored if our principles had been utilized. First, because Mark made preparations to care for Michael during the period between conception and 90 days after the infant's birth, he would automatically be considered a presumed parent. This issue resolved, the case could be immediately heard within a month. The issues of standing, fitness, and custody would be heard in a combined hearing, and any appeals would be addressed in the same expedited manner. Attachment and other issues involved in the child's best interest would be identified by the specially-educated judges and the professional child development mediators. Because of the stringent time frame and new respect for the personhood of the infant, Michael's interests would be served no matter what the outcome of the trial. Whether he be left in the custody of the adoptive parents or given over to his father, he would have the opportunity to develop secure, loving relationships with his caregivers.
*Professor of Psychology, Santa Clara
**Student, Professional School of
Psychology, Fresno, California. The authors wish to acknowledge the collaboration
of Katherine Hohn (J.D. 1995, Santa Clara University).
1Adoption of Michael H., 29 Cal.
Rptr. 2d 251 (Ct. App. 1994), rev'd, 898 P.2d 891 (Cal. 1995). The facts
set forth below are contained in the court's reported decision.
21 John Bowlby, Attachment and Loss
3Robert R. Sears et al., Patterns
of Child Rearing (1957).
4Jerry Burger, Personality (2nd ed.
5John Bowlby, Child Care and the
Growth of Love (Mary D. Salter ed., 2d ed. 1965).
6Calvin S. Hall et al., Introduction
to Theories of Personality (1985).
7Mary S. Ainsworth & Barbara A. Wittig,
Attachment and Exploratory Behavior of One-Year-Olds in a Strange Situation,
in 4 Determinants of Infant Behavior 111, (Brian M. Foss ed., 1969).
8Charles Darwin, The Expression of
Evolution in Man and Animals (1872).
9See Bowlby, supra note 2.
10Carol George et al., The Adult
Attachment Interview (1985) (on file with University of California, Department
of Psychology, Berkeley).
11Mary S. Ainsworth et al., Patterns
of Attachment: A Psychological Study of the Strange Situation (1978).
12A fourth pattern of attachment
called disorganized/disoriented insure attachment has been recognized
by developmental psychologists. Due to its recent identification and primarily
clinical use, however, the authors have chosen not to include it for further
discussion in this paper.
13L. Alan Sroufe, Relationships,
Self, and Individual Adaptation, in Relationship Disturbances in Early
Childhood 70 (Arnold J. Sameroff & Robert N. Emde eds., 1989).
14Ainsworth et al., supra note 11.
15Elizabeth Anisfeld et al., Does
Infant Carrying Promote Attachment? An Experimental Study of the Effects
of Increased Physical Contact on the Development of Attachment, 61 Child
Dev. 1617, 1617-27 (1990).
16Ainsworth & Wittig, supra note
17Marc H. Bornstein & Catherine
S. Tamis-LeMonda, Activities and Interactions of Mothers and Their Firstborn
Infants in the First Six Months of Life: Covariation, Stability, Continuity,
Correspondence, and Prediction, 61 Child Dev. 1206 (1990).
18Everett Waters et al., Attachment,
Positive Affect, and Competence in the Peer Group: Two Studies in Construct
Validation, 50 Child Dev. 821 (1979).
19Margaret S. Mahler et al., The
Psychological Birth of the Human Infant: Symbiosis and Individuation (1975).
20Jerome Kagan, The Second Year:
The Emergence of Self Awareness (1981).
21G.J. Suess et al., Effects of
Infant Attachment to Mother and Father on Quality of Adaptation in Preschool:
From Dyadic to Individual Organization of Self, 15 Int'l J. Behavioural
Dev. 43, 43-65 (1992).
22Leah Matas et al., Continuity
of Adaptation in the Second Year: The Relationship Between Quality of
Attachment and Later Competence, 49 Child Dev. 547 (1978).
23Adriana G. Bus & Marinus H. Van
Ijzendoorn, Attachment and Early Reading: A Longitudinal Study, 149 J.
Genetic Psychol. 199 (1987).
24Michael Tomasello & Michael J.
Farrar, Joint Attention and Early Language, 57 Child Dev. 1454 (1986).
25L. Allen Sroufe et al., Individuals
in Relationships: Development from Infancy Through Adolescence, in Studying
Lives Through Time: Personality and Development (David C. Funder et al.
26Mary Main & Jude Cassidy, Categories
of Response to Reunion with the Parent at Age Six: Predictable from Infant
Attachment Classifications and Stable over a One Month Period, 24 Developmental
Psychol. 415 (1988).
27Sroufe et al., supra note 25.
29Mary S. Ainsworth, Attachments
Beyond Infancy, 44 Am. Psychologist 709 (1989).
30Sroufe et al., supra note 25.
31Sroufe et al., supra note 25;
R. Rogers Kobak & Amy Sceery, Attachment in Late Adolescence: Working
Models, Affect Regulation, and Representations of Self and Others, 59
Child Dev. 135 (1988).
32George et al., supra note 10;
Malcom L. West & Adrienne E. Sheldon-Keller, Patterns of Relating (1994);
Cindy Hazan & Phillip Shaver, Romantic Love Conceptualized as an Attachment
Process, 52 J. Personality & Soc. Psychol. 511 (1987) [hereinafter Romantic
Love]; Cindy Hazan & Phillip Shaver, Love and Work: An Attachment-Theoretical
Perspective, 59 J. Personality & Soc. Psychol. 270 (1990) [hereinafter
Love & Work].
33West & Sheldon-Keller, supra note
34Ainsworth & Blehar, supra note
35Romantic Love, supra note 32.
36R. DeBlasio & Eleanor W. Willemsen,
Attachment Security and Relationship Satisfaction (paper presented at
the Western Psychological Association Conference, May 1994, Kona, HI.)
(on file with author).
37Peter Fonagy et al., Maternal
Representations of Attachment During Pregnancy Predicts the Organization
of Infant-Mother Attachment at One Year of Age, 62 Child Dev. 891 (1991).
38Jeffrey A. Simpson et al., Support-Seeking
and Support-Giving Within Couple Members in an Anxiety-Provoking Situation:
The Role of Attachment Styles, 62 J. Personality & Soc. Psychol. 434 (1992).
39Judith A. Crowell & Shirley S.
Feldman, Mothers' Working Models of Attachment Relationships and Mother
and Child Behavior During Separation and Reunion, 27 Developmental Psychol.
40Patricia M. Crittenden et al.,
Family Patterns of Relationship in Normative and Dysfunctional Families,
3 Dev. & Psychopathology 491 (1991); Charles H. Zeanah & Paula D. Zeanah,
Intergenerational Transmission of Maltreatment: Insights from Attachment
Theory and Research, 52 Psychiatry 177 (1989); Mary Main & Ruth Goldwyn,
Predicting Rejection of Her Infant from Mother's Representation of Her
Own Experience: Implications for the Abused-Abusing Intergenerational
Cycle, 8 Child Abuse and Neglect 203 (1984).
41Judith A. Feeney et al., Assessing
Adult Attachment, in Attachment in Adults: Clinical and Developmental
Perspectives 128 (Michael B. Sperling & William H. Berman eds., 1994).
42 Mario Mikulincer & Orna Nachshon,
Attachment Styles and Patterns of Self-Disclosure, 61 J. of Personality
& Soc. Psychol. 321 (1991).
43Nancy L. Collins & Stephen J.
Read, Cognitive Representations of Adult Attachment: The Structure and
Function of Working Models, in 5 Advances in Personal Relationships 53
(Daniel Perlman & Kim Bartholomew eds., 1994).
44Love & Work, supra note 32.
45 J.C. Rothbard & Phillip R. Shaver
(1994), Continuity of Attachment Across the Life Span, in Attachment in
Adults: Clinical and Developmental Perspectives (M.B. Sperling & W.H.
Berman eds., 1994); Deborah B. Jacobvitz et al., The Transmission of Mother-Child
Boundary Disturbances Across Three Generations, 3 Dev. & Psychopathology
513 (1991); Zeanah & Zeanah, supra note 40; Marinus H. Van Ijzendoorn,
Intergenerational Transmission of Parenting: A Review of Studies in Nonclinical
Populations, 12 Developmental Rev. 76 (1992); Main & Goldwyn, supra note
46Suess et al., supra note 21; Frits
A. Goossens & Marinus H. Van Ijzendoorn, Quality of Infants' Attachments
to Professional Caregivers: Relation to Infant-Parent Attachment and Day
Care Characteristics, 61 Child Dev. 832 (1990); Susan Kromelow et al.,
The Role of the Father in the Development of Stranger Sociability During
the Second Year, 60 Am. J. Orthopsychiatry 521 (1990).
47Clair B. Kopp, Trends and Directions
in Studies of Developmental Risk, in Threats to Optimal Development: Integrating
Biological, Psychological, and Social Risk Factors (Charles A. Nelson