This excerpt from Attachment and Bonding A New Synthesis C. Sue Carter, Lieselotte Ahnert, K. E. Grossmann, Sarah B. Hrdy, Michael E.

Transcription

1 This excerpt from Attachment and Bonding A New Synthesis C. Sue Carter, Lieselotte Ahnert, K. E. Grossmann, Sarah B. Hrdy, Michael E. Lamb, Stephen W. Porges and Norbert Sachser 2006 The MIT Press. is provided in screen-viewable form for personal use only by members of MIT CogNet. Unauthorized use or dissemination of this information is expressly forbidden. If you have any questions about this material, please contact cognetadmin@cognet.mit.edu.

2 10 Universality of Human Social Attachment as an Adaptive Process K. E. GROSSMANN andk. GROSSMANN Institut für Psychologie, Universität Regensburg, Regensburg, Germany ABSTRACT Attachment is the phylogenetically programmed propensity of a human child to form a special relationship with responsive caregivers as part of the infant parent bond. Attachment theory is rooted in traditional ethology as well as in psychoanalysis. Infants actively seeking protection and care has most likely been one of the selective forces. Other bonds between individuals may contain certain elements of attachment. During ontogenesis, individual differences in the organization of emotions and integration into a coherent internal working model throughout the years of immaturity is particularly prominent in human individual development. Secure and insecure patterns of attachment result from different qualities of interaction between infants expressions and adults responses to them. In adulthood, secure attachment representations reflect a sense of being worthy of help, free access to supportive attachment figures when needed, and appropriate, partnership-oriented evaluations of challenging social situations. Secure internal working models constantly integrate coherently and adaptively emotional and relevant cognitive aspects across lifetime. Secure attachment and secure exploration serve psychological security and quality of adaptation to life in the mental health sense. A longitudinal study conducted in northern Germany demonstrates that children s experience with both parents influenced their psychological security in representation of attachment as well as partnership representation at age 22 years. Currently, traditional attachment theory is on its way to being integrated into modern sociobiology and physiology. A future synthesis of attachment and bonding could benefit from addressing Tinbergen s four questions about the evolution, function, causation, and development of attachment and bonding. ATTACHMENT, BONDING, AND RELATED ISSUES Historical Notes Attachment theory, as conceptualized by John Bowlby (1987), focuses on the nature of the child s tie to the mother. It posits a pre-programmed propensity in

3 200 K. E. Grossmann and K. Grossmann the young child to develop attachments to a very few, special adults who are willing and able to care for, support, and invest resources in the young individual. It was designed to serve as a theoretical concept for human emotional and social development and to explain qualitative individual differences in the ontogenesis of psychological adaptation and mental health, depending on the quality of care received. Stated within the framework of modern sociobiology that was unknown to Bowlby, attachment theory focuses on the gene-selfish interest of the child to receive as much of his parents (physical as well as psychological) resources as needed to become an emotionally and socially healthy and competent individual in his social group. In terms of Trivers (1974) parent offspring conflict, attachment theory focuses on the offspring s side of that conflict as well as on the parent s willingness or unwillingness to invest in this individual offspring. In the case of parental unwillingness, infant survival is at stake. For example, among the langurs of Abu, allomothering is abundant, but no mother was ever seen to allow any infant but her own to suckle (Hrdy 1977, p. 211). Therefore, it is vitally important for an infant to be able to identify and seek out its mother extremely soon; otherwise, it will find no other source of nourishment. In support of this assumption, Hrdy observes that an infant langur knows its mother very well despite extant allomothering. An infant that has been quietly holding its mother may begin to whine soon after being taken by some other female. Conversely, an infant that has been struggling and complaining will usually grow quiet when retrieved by the mother (Hrdy, 1977, p. 215). From the maternal perspective in this conflict, rejecting an infant carries a much lesser cost to the mother, as she can expect another offspring. In four interdisciplinary meetings of the World Health Organization Study Group on the Psychobiological Development of the Child (Geneva 1953, 1955, 1956 and London 1954) John Bowlby encountered ethology of the 1950s as Lorenz (1960) and Tinbergen (1951) formulated it (Tanner and Inhelder 1960). Bowlby was tutored in this new discipline by the comparative biologist Robert Hinde (Bowlby 1991; Hinde 2005). Bowlby aimed at providing a theoretical frame for a new synthesis of phylogeny, ontogeny, psychoanalytic insights, inner control systems (e.g., schemata, equilibration sensu Piaget; Bowlby 1960) and Darwinian thinking. Attachment theory was formulated by Bowlby to encourage empirical examination of the right questions of psychoanalysis to which the psychoanalysts, as he believed, often gave the wrong answers according to modern scientific standards. He perceived a need to study prospectively healthy as well as maladaptive social-emotional development of individuals from the cradle to the grave (Bowlby 1979). One of the central observations that paved the way for the concept of behavioral systems in attachment theory was the behavioral response patterns demonstrated by Harlow s rhesus monkeys, which were dependent on their emotional state. In the presence of their surrogate mother and when calm, infant monkeys would play freely in an open field. However, when alarmed and frightened, they

4 Universality of Human Social Attachment 201 did not run away from the frightening object but rather fled to their surrogate mother figure as a safe haven, where they could calm down. In the absence of the mother figure, no calming down was observable. Other vivid examples of the functioning of the aroused attachment system in mammals have been provided by Robert B. Cairns (1979, p. 47), who described the abrupt transformation of a one-month-old monkey from a relaxed, content infant into a restless, crying, and sometimes extremely agitated organism through forced separation from its mother. Lambs separated from other sheep run full speed around the isolation chamber, sometimes against the door or wall; this behavior is repeated with groggy redundancy. The co-regulation of the attachment need of the young and its corresponding maternal caregiving system was termed affectionate system by Harry Harlow (1961, 1971). The behavioral systems are fitted to each other and have been phylogenetically selected as a unit of age-specific adaptive behaviors of infants and mothers if, as modern sociobiology would suggest, the mother has decided to invest in this infant. The child s need is expressed through attachment behaviors accompanied by expressions of emotions (Darwin 1872/1998), which function to increase close proximity to the caregiver. The caregiving behavioral system is expressed by protecting and caring for the attached individual. The need for care and protection of a dependent or weaker individual which may also be an adult in times of ill health, adversity, or old age by a stronger and wiser adult is central to Bowlby s conceptualization. Because care and protection are vital needs of weaker and inexperienced younger individuals, the attachment behavioral system is driven by strong motivations. Many of the most intense emotions arise during the formation, the maintenance, the disruption and the renewal of attachment relationships (Bowlby 1979, Lecture 7). This concept was certainly influenced by Bowlby s training as a psychotherapist, as well as by having seen the misery of many children throughout Europe who were separated from or had lost their parents after World War II (Bowlby 1951). Of course, human beings also form affectional bonds with others in the course of their life. Ainsworth (1985, p. 199) highlighted the difference between attachment of infants to their mothers and affectional bonds in other social relations as follows: An attachment is an affectional bond in which the attachment figures are never wholly interchangeable with or replaceable by another, even though there be another to whom one is also attached. In addition, she refers to developmental changes in the way in which a child s attachment to parents manifests itself beyond the infancy period, that one s attachment to parents tends to persist throughout life rather than attenuating and eventually disappearing as many believe. In comparison, bonds are also relatively long-lived ties to unique individuals.a bond is to be distinguished from other long-term relationships (attachments) in which it is the role of the other that is significant, so that with

5 202 K. E. Grossmann and K. Grossmann separation or loss there would be at most some regret, tempered by an expectation of soon finding another to play the same role in one s life (Ainsworth 1985, p. 799). Later, Ainsworth adds, There is one criterion of attachment that is not necessarily present in other affectional bonds. This is a seeking to obtain an experience of security and comfort in the relationship with the partner. If and when such security and comfort is available, the individual is able to move off from the secure base provided by the partner, with confidence to engage in other activities (Ainsworth 1991, p. 38). Among those other affectional bonds throughout the life span she lists the bond of mother to infant (as contrasted to the phylogenetically programmed process of infants forming attachments to individual caregivers), the bond of father to child, bonds resulting from sexual behavior, friendships, companionships, as well as bonds between siblings and other kin. All of these bonds may develop into attachment relationships, and they appear to be somehow influenced by attachment experiences. These other relationships, however, are characterized by properties of their own, which may be well beyond attachment (Hinde 1976; Dunn 1993). We view attachment as the developmental process during which infants genetic programs become phenotypically manifest, observable, and testable as a function of caretakers responsiveness. Historically, the concept of bonding entered developmental psychology through the work of Klaus and Kennell (1976). They argued for an imprinting type of maternal readiness to establish a close relationship to the infant in the human mother, as in ungulates, that has to be triggered within the first hour(s) after delivery. The process by which a mother comes to feel close to her infant has been termed bonding in the developmental psychology literature ever since. We have tested this assumption by providing a group of 24 German mothers with early contact for the first hour after birth and compared their tender behavior toward their newborn with 25 mothers who did not receive early contact. Results showed that non-early contact mothers reached the level of tenderness of the early contact mothers by the second week, and that later quality of maternal sensitivity or infant mother attachment was not affected by our experimental manipulation of maternal bonding (Grossmann, K. et al. 1981). Below, we present our arguments on the universality of human social attachment as an adaptive process using the following framework: 1. Development of the attachment system in human infants. 2. Different qualities of secure and insecure attachment patterns. 3. Attachment development from infancy to adulthood. 4. Interrelations between sociocultural and cognititive development within attachment development as the infant learns about the motivations, intentions, and perspectives of their mothers and incorporates them naturally into their own cultural world views. In the second section, we present the idea, first developed by Lorenz (1967), that attachment is a universal genetic program valid in all cultures, despite clearly

6 Universality of Human Social Attachment 203 observable variations in parental caregiving behaviors between existing cultures and within cultures in different epochs. This will be augmented by results of studies using the separation reunion procedure with infants and their mothers in various cultures that support the universality claim of attachment theory. In the third section, some of the primary results from our German longitudinal studies are discussed as an example for variations in attachment development in a group of children growing up in their families of origin. Finally, new theoretical considerations are presented that appear to be important for a new synthesis of attachment and bonding. Development of the Attachment System Although human infants are physically premature at birth, they are equipped with communicative competencies inherent in the child s emotional expressions and behaviors; under most circumstances these are recognized by the caregiver as bodily and social needs and are responded to appropriately (Ainsworth and Bell 1974; Trevarthen, this volume). The infant is not born with an attachment to his mother; this begins to develop after birth. Still, development of an attachment is pre-programmed such that given a caregiver who is reliably available over a long period, an infant will necessarily develop an attachment to that particular adult independent of the quality of caregiving, even to a nonresponsive surrogate mother (Harlow 1971) or abusive mother (Hennighausen and Lyons- Ruth, this volume). The quality of care, however, will have a major influence on certain aspects of the personality development of that infant and child (e.g., how behaviors and intentions of others are perceived and responded to). Ainsworth and her team spent many hours observing infants and mothers in their natural environment. Their results showed that prompt and appropriate responsiveness to all the infant s behaviors shaped the infant s attachment development toward a secure individual attachment from early on, whereas unresponsive, uncooperative, inappropriate care paved the way for an insecure attachment. Mothers who responded promptly and sensitively to their infants crying and other emotional expressions more frequently experienced infants who cried less toward the end of the first year and who were more compliant with their mother. Likewise, giving the baby close bodily contact when he signaled for it was associated both with secure attachment and the growth of self-reliance (Ainsworth et al. 1974; Grossmann, K. et al. 1985). In contrast to the prevailing learning theory at that time, Ainsworth and Bell (1977) offered an ethological interpretation, which corresponded to a shift toward a new phylogenetic paradigm: If all infant behaviors are seen as acts of communication instead of accidental operant behaviors, then qualitatively different responsiveness by the adult will lead to qualitatively different communicative competencies and developmental consequences for the child rather than just changing the frequency of certain specific behaviors (Ainsworth and Bell 1974).

7 204 K. E. Grossmann and K. Grossmann Individual Qualities or Patterns of Attachment of Mother Child Dyads Ainsworth tested the validity of her home observations in the so-called strange situation (Ainsworth and Wittig 1969). Because an infant will not show attachment behaviors under conditions of well-being, Ainsworth devised a standardized procedure to arouse the infant s attachment system. The Strange Situation assessment of attachment relationship emphasizes the infant s responses to mild stress induced by a novel environment, novel person, and two brief separations from the mother. During separations from the attachment figure, infants usually limit or stop their exploratory activities (later we will point to the universality of this phenomenon). A pattern of attachment behaviors throughout the Strange Situation was considered secure if the attachment person functioned well as a haven of safety calming the infant and alleviating anxiety and distress, and as a secure base from which the infant resumed exploratory and playful orientation to the world beyond the attachment figure after the aroused attachment system had been calmed (see Figure 10.1). Attachment behavior patterns were considered insecure if an infant showed behavioral restrictions or limitations when using the mother as a safe haven by not seeking bodily proximity to her or when using the mother as a secure base by not wanting to move off and explore again after separation. Subsequently, three major patterns of an infant s behavioral strategy in response to separation from and reunion with the attachment figure were distinguished: secure, insecure-avoidant, and insecure-resistant. The most general marker of the three patterns is appropriate expression, hypo- or hyperactivation of attachment relevant behaviors. Ainsworth s home studies suggested that infants patterns of attachment in the Strange Situation procedure were a consequence of variations in sensitivity of maternal responsiveness to their baby s attachment signals and their signals of wanting to explore during the first year. In conditions of emotional distress: Attachment behavior In conditions of emotional well being: Exploratory behavior Attachment behavior Figure 10.1 The concept of the attachment exploration balance. The attachment and exploratory behavioral systems are linked to each other. When one system is activated the other is deactivated.

8 Universality of Human Social Attachment 205 Studies of attachment development, spanning from preschool age to old age, included the symbolic and representational level of perceiving, feeling, thinking, and discussing attachment. Secure and insecure patterns of behaving or thinking were reliably found in response to the aroused attachment system and in relation to attachment figures (see Cassidy and Shaver 1999). Individuals responding with a secure strategy to distress show a free, uninhibited access to an attachment figure when alarmed and explore readily when the attachment figure is (mentally) present and responsive. The studies neither under- nor overemphasize attachment behaviors, exploration, or their mental representations. The patterns can be shown in narrations, reflecting an internalization of such processes. The secure pattern has been considered by Ainsworth as psychologically more adaptive because reliance on the trusting support of significant others increases the chances of mastering new challenges. The three major patterns of attachment have been confirmed and validated in many studies across many countries (van IJzendoorn and Kroonenberg 1988; van IJzendoorn and Sagi 1999). In addition, Main and Solomon (1986) described the behavior of infants who either could not be classified or who showed signs of extreme disturbance such as fear of the attachment figure. They were classified as insecure disorganized and disoriented in their attachment. Evidence for attachment insecurity of this group came from an analysis of the cortisol level of these infants. Spangler and Grossmann (1993) showed that all infants with signs of disorganization had elevated cortisol levels after the Strange Situation as compared to infants with a secure pattern of attachment that showed no increase in cortisol level. Disorganization in response to distress plays an increasingly important role in clinical and otherwise psychologically maladapted samples (Hennighausen and Lyons-Ruth, this volume). Attachment across the Life Course Attachment development has been conceptualized as a continuous process not only in infancy, but well into adolescence. Principal determinants of the pathway along which an individual s attachment behavior develops, and of the pattern in which it becomes organized, are the experiences he has with his attachment figures during his years of immaturity infancy, childhood and adolescence (Bowlby 1980, p. 41). This quotation cautions those who perceive attachment theory and research as relating only to infancy and early childhood. Attachment theory posits instead that attachment experiences result in internal working models (IWMs) or expectations, which begin from an early age: how the attachment figures will respond when the person finds herself in conditions of adversity which evoke feelings of anger, fear, and sadness. IWMs of attachment are thought of as a set of rules and filters regulating the perception of the social world. Working models of attachment can be secure or confident to be worthy of help, insecure and dismissing of the importance of attachment, or insecure and enmeshed/confused about the meaning of attachment experiences.

9 206 K. E. Grossmann and K. Grossmann IWMs of attachment are assessed through interviews about current thoughts on the person s own attachment experiences in childhood and adolescence (Hesse 1999). Longitudinal studies in attachment development spanning from infancy to young adulthood, such as our Bielefeld and Regensburg studies (Grossmann, K.E. et al. 2005), the Minnesota study (Sroufe et al. 2005), and the Berkeley study (Main et al. 2005), leave no doubt that there are many long-lasting influences of attachment experiences with parents in childhood and adolescence on social and personality development (Thompson 2000; Waters and Cummings 2000). During attachment development, sensitivity to the infant s attachment signals is important, as are sensitivity to exploratory intentions and behaviors that support a balanced organization of behavior. According to Konrad Lorenz (1967, p. 383), exploration means Forschung (i.e., research), which needs a relaxed atmosphere free from anxiety or other pressures. Exploration is necessary for adaptation to the wider social and material world as new and increasingly complex experiences have to be integrated throughout development. The indispensable role of exploration for healthy personality development has been part of attachment theory right from the onset (Bowlby 1979, p. 133). Playful exploration of the physical and mainly cultural world at large and the people in it profits from a secure base or a secure IWMof attachment that needs to be constantly updated during development to function adaptively. The markers of psychological security, as described by most attachment researchers, are physical or mental access to others in times of distress as well as the ability to clarify and manage complex and unhappy social emotional challenges by developing realistic solutions, including open communication with others and seeking and accepting their help; in short, the capacity to overcome anxiety and fear through proximity and with the help of others. It should be noted, however, that without the arousal of the attachment system, everyday routine problem solving will prevail. Psychological security is most relevant in emotional testing the limits situations, which require updates of IWMs as in psychological adaptation in the elderly (Kliegl et al. 1989). Attachment relationships as the first and often only relationships in infancy are pivotal to an individual s social relationships and their evaluations, to partnership-oriented planning, and particularly to finding goals worth pursuing (Minsky 1987, p. 175). Later attachment relationships and IWMs of attachment seem to influence the person s view about the world, whether perceived as being threatening or benevolent. This has been shown for all age periods from childhood (Suess et al. 1992) to old age (Wensauer and Grossmann 1998). Quality of attachment and quality of support during exploration, in our view, function jointly as a new system with novel and special qualities resulting in psychological security or insecurity. The quality of psychological security organizes and integrates emotions, cognition, and communicative discourse for intricate adaptations to special other people.

10 Universality of Human Social Attachment 207 In summary, attachment refers primarily to the young child s need for protection, care, and support for adaptation to his social and nonsocial environment. Attachments are the natural prerequisite for becoming emotionally and socially acculturated. Attachment relationships set the stage for cognitive interpretation of experiences and for acquiring meaning about self and others in complex social situations that are vital in infancy but continue to be important throughout life. Attachment and Sociocultural Cognition The acquisition of meaning relies on more knowledgeable persons providing an interpretation of the world within a child s context of immediate action and experience. Language, as the obvious carrier of meaning, is learned and functions best in situations of joint attention of young and older humans (Tomasello 1999). Joint attention for the sake of acculturation is prominent only in the human species. In situations of joint experiences, infants approaching their first birthday come to be aware of the other s mind. The adult partner will most certainly be an attachment figure or another familiar and trusted person, because, by that age, fear of strangers will be functioning strongly. It seems as if nature wanted to make sure that infants begin their lifelong education by first learning about the values of their own people. The subsequent stages of joint attention, following and guiding attention of the adult partner, appear around 11 to 13 months of age, closely linked to language development. These might already be shared with more interactive partners. In any case, Forschung in exploration and in joint attention is inseparably linked to cognitive as well as emotional development. In addition, discourses during joint attention and narratives about joint and observed experiences help to link inner states and feelings cognitively to external events (Grossmann, K.E. 1999; Nelson 1999). According to Bowlby (1982, p. 355), by sharing a common set-goal and participating in a joint plan to achieve it, partners have a rewarding sense of common purpose; and they are likely also to identify with one another. From a cognitive perspective, psychologically adaptive or secure behavior strategies in face of adversity are the result of undistorted perceptions of relevant situations that demand well planned, goal-corrected actions. Robert Sternberg (1997), an intelligence researcher, sees adaptive intelligence as a cognitive correspondence between a complex outside reality and its coherent representation in the human mind ( internal coherence and external correspondence ). In attachment theory, reality often refers to close other people s minds, feelings and intentions, and internal coherence to the organization of one s emotions as well as one s knowledgeable insights and reflections. Sternberg s concept of adaptive intelligence helps us understand how the historical cultural dimension, with its cultural artifacts and its culture-specific meanings, pre-programmed during evolution, is transmitted during ontogenesis. The overall

11 208 K. E. Grossmann and K. Grossmann model is thus that human beings have cognitive skills that result from biological inheritance working in phylogenetic time; they use these skills to exploit cultural resources that have evolved over historical time; and they do this during ontogenetic time (Tomasello 1999, p. 48). Strong and emotional identification with the values of one s culture requires persons that are respected and loved by the child (Lorenz 1967, 1977); it needs attachment learning (Minsky 1987), meaning conveyed through narrative discourses (Bruner 1990), social mediation (Cole 1996), and community participation (Rogoff 2003). The emotional organization established in the early years seems to be a basic component of an adaptive synthesis of the cognitive system and the emotional system especially in challenging situations (Grossmann, K.E et al. 2002; Sroufe 2000). Man s capacity to use language and other symbols, his capacities to plan and build models, his capacities for long-lasting collaboration with others and for interminable strife, these make man what he is. All these processes have their origin during the first three years of life, and all, moreover, are from their earliest days enlisted in the organization of attachment behavior (Bowlby 1982, p. 358). EVIDENCE FOR THE UNIVERSAL NATURE OF ATTACHMENT Relevant specific endowments for attachment as well as cultural development in humans are seen in a number of behavioral and mental programs: the organization of emotions (Cosmides and Tooby 2000), joint attention (Tomasello 1999), language and the creation of meaning (Bruner 1982, 1983, 1990), awareness of other s minds including the ability to see events from the perspective of another person (Ainsworth et al. 1974; Meins 1999), the ability to consider another person s wishes and goals when pursuing own goals ( goal-corrected partnership, Bowlby 1982), self-cognition (Cassidy 1988), and autobiographical narrative constructions (Nelson 1996, 1999). With development, these universal mental and representational programs become increasingly individualized and differentiated in the course of interactions with special individuals, who are the transmitters of cultural meaning. Nature has programmed human beings as cultural beings. Cultural learning begins within the attachment relationship. Attachment is not an option with which culture can freely play. It is a universal necessity for individual human development. History has shown that there are definite limits of human adaptability to inhumane child-rearing conditions that disregard attachment. Examples can be seen in the legendary sad outcomes of infants raised in large de-individualized groups, as described by René Spitz (1945), or in the German

12 Universality of Human Social Attachment 209 Nazi Project Lebensborn, a project intended to produce a purely Aryan race but which removed infants from their mothers (Lilienthal 1993; Abe 2003), or, more recently, in Romanian orphanages (O Connor, this volume). The following concepts may serve to exemplify the universal nature of attachment. First, the universality concept of attachment states that all infants become attached to at least one primary caregiver. Infants and young children may become attached to more than one person, such as fathers and regular caregivers (Sagi et al. 1995), but not to many. As regular interactive partners, additional familiar persons may become secondary attachment figures in the hierarchy of attachment figures. Although the attachment system operates independently of genetic kinship, the biological mother is usually the primary attachment figure as she is typically the primary caregiver. Sleeping together may be an important experience in attachment development. A survey of 90 non-western cultures revealed that, in every one, infants and toddlers slept in the same bed with the mother or a closely related adult (Barry and Paxson 1971). By comparison, a study of Israeli infants showed that the majority of infants sleeping in community houses, instead of their parents home, showed disorganized attachment behaviors (Sagi et al. 1994). This was also found for infants of divorced parents if, as a result of joint custody, the infant slept alternately at the mother s and father s homes (Solomon and George 1999). Affectionate responsiveness, attentive protection, sensitive support, and challenges during joint play seem to suffice for attachment formation and become highly influential across the life span (Grossmann, K.E. et al. 2002; Thompson 2000). A person has become an attachment figure to a child if three universal criteria are met: (a) the young child seeks closeness to and protection from that person when it perceives danger or experiences distressed (the haven of safety function); (b) the young child uses that person as a secure base from which it explores the environment, and (c) the young child will experience emotional as well as physiological distress when separated from that figure (separation distress). By the end of the second year, the toddlers of our Bielefeld longitudinal study had up to three attachment figures, comparable to the toddlers in the Glasgow study (Schaffer and Emerson 1964). Next, the concept of the attachment exploration balance implies that when the attachment behavioral system is active or aroused, the exploratory behavioral system is less active and vice versa (see Figure 10.1). This activation and deactivation does not to seem be complete, however, to the degree of closing down one or the other system completely. Environmental clues to danger are continuously monitored as are events stimulating curiosity. Arousal of the attachment system is best calmed by close bodily proximity ( tender loving care ). Nonarousal of the attachment system can be taken as an index of emotional well-being, which favors playful exploration in the service of competent social and cultural adaptation to the complexities of human life. The universal functioning of the attachment exploration balance is experienced most likely

13 210 K. E. Grossmann and K. Grossmann by anyone who comes too close to an unfamiliar toddler in any culture. The toddler will often flee in panic to her mother or other familiar figure. This functioning has been systematically and empirically demonstrated in many cultures (van IJzendoorn and Sagi 1999). We inspected and collected data gained from Strange Situation studies from four very different cultures. In each group of toddlers, exploratory activity declined markedly when a stranger entered the room and wanted to interact with the toddler and separation from the mother was enforced. This decline in exploration with increasing distress was shown by toddlers in Baltimore, U.S.A. (Ainsworth et al. 1978), by German toddlers in various cities (Grossmann, K. et al. 1985; Grossmann, K.E. 1999), by Japanese toddlers (Miyake et al. 1985), and by Trobriand toddlers (Grossmann, K.E. et al. 2005). Exploration was dominant when the U.S./German/Japanese/Trobriand toddler was with her mother; the entrance of a stranger reduced exploration to a certain degree, and it declined further or remained at a low level after the mother departed. The mother s return resulted in some increase of exploratory activity, although not nearly to the level that was observed prior to separation. Thus, even short separations (maximum of 3 min) from the secure base made all toddlers wary or anxious. Separations from the mother had a lasting dampening effect on infant exploratory activity in a strange environment in all of the observed toddlers from four continents, regardless of their pattern of attachment. Third, the normativity hypothesis states that the secure pattern of attachment is with very few exceptions the most frequent pattern in all samples of uncompromised infants observed in many cultures. Ainsworth (1979, p. 44) proposed that the secure pattern is to be considered as normative and the optimal adaptive developmental outcome for human infants. Subsequently, there have been many studies of infant patterns of attachment to mother, father, and sometimes to a permanent nonparental caregiver. The Strange Situation procedure has been conducted with infants from many cultural backgrounds in the Western and Eastern hemispheres, from agriculturally oriented to industrialized peoples. Almost all confirmed this proposition (van IJzendoorn and Sagi 1999). Only two exceptions have been published: (a) in our northern German sample, but not in several southern German samples, the insecure-avoidant pattern was found to be dominant (Grossmann, K. et al. 1985); (b) in an Israeli sample of infants with out-of-home sleeping arrangements, the insecure-ambivalent pattern and even disorganized attachment behaviors were dominant (Sagi et al. 1994). Finally, the sensitivity hypothesis states that the quality of attachment depends on the attachment figure s sensitivity to the child s attachment and exploratory behaviors, that is, the quality of care received. Quality of care even extends to being given enough food under poor economic living conditions (Valenzuela 1997). Secure or insecure patterns of attachment are not genetically determined, as qualities of attachment to two or even three attachment figures are largely independent of each other (van IJzendoorn et al. 1992). Studies investigating

14 Universality of Human Social Attachment 211 antecedents of security of attachment conducted in the U.S.A., Europe, and in other cultures all demonstrated a significant relationship between maternal sensitivity and a secure pattern of infant attachment. Although each research team defined maternal sensitivity in a somewhat different way, no study yielded contradictory evidence (De Wolff and van IJzendoorn 1997). Some cultural institutions, however, clearly hinder parents in providing sensitive and cooperative care. General cultural beliefs that crying strengthens the lungs or responding to crying spoils a child and creates a tyrant in the house have dominated child-rearing customs in Germany in the past and have kept many a mother from responding to her crying infant (Grossmann, K. et al. 1985). Some social groups may even limit parents access to their children, as formerly in some traditional Israeli kibbutzim (Aviezer and Sagi 1999). DEVELOPMENT OF PSYCHOLOGICAL SECURITY: A NORTHERN GERMAN LONGITUDINAL STUDY OF ATTACHMENT DEVELOPMENT IN THE FAMILY Attachment theory posits a causal relationship between an individual s experience with his/her parents and his/her later capacity to make affectional bonds. Enduring and supportive bonds between adult partners, together with the capacity to care and make affectional bonds with one s own infants, are optimal preconditions for a joint and successful upbringing of one s offspring. This is interpreted as the adaptive value of attachment and sensitive caregiving. Evidence supporting this hypothesis can be found within one of our longitudinal studies of a group of 38 young German adults, whom we observed since birth as they grew up in their nonrisk families of origin. At 22 years of age, the young adults representations of partnership, including markers of attachment, were assessed based on a German adaptation of the current relationship interview by Crowell and Owens (see Crowell and Waters 2005). A secure partnership representation describes a person who values an intimate relationship as a reliable source of comfort, who has a mutual secure-base relationship with the partner, and who has an easy access to partnership-related thoughts and feelings without denying unfavorable aspects of that partnership. Trained evaluators rated security versus insecurity of partnership representation on one dimension and quality of discourse about partnership on another. Quality of discourse was an important marker of the capacity to make affectional bonds for those young adults who did not experience a mutually supportive partnership but could still present a clear and coherent discourse about his/her disappointing partnership experiences, be open and competent in his/her discourse about partnership and address freely attachment aspects within a partnership. Security and quality of discourse were correlated significantly (r = 0.42). Longitudinal analyses revealed three major influences on the quality of partnership representation in childhood experiences with both parents starting from

15 212 K. E. Grossmann and K. Grossmann the early years: (a) maternal sensitivity, (b) the child s quality of discourse when interviewed about attachment and friendship issues, and (c) the quality of father s sensitive and challenging interactions during play with his toddler (Grossmann, K. et al. 2005; Grossmann, K.E. et al. 2002). Maternal sensitivity was observed and assessed in situations appropriate to the child s development during the first year, at 24 months, and at age six. Later, mothers and children/adolescents were interviewed, and maternal support, acceptance, and empathic understanding of the child, as reported by each side, was rated. Aggregation of maternal sensitivity ratings with ratings of maternal support for each age of assessment yielded aggregated scores for the age periods infancy, childhood, and adolescence. These scores revealed a significant influence for each aggregated score on security of partnership representation and quality of discourse about partnership. In particular and surprisingly, a significant relationship emerged between observed maternal sensitivity and support for exploration during the infancy years and quality of discourse about partnership. We interpret these results as reinforcing the assumption that interactive experiences with the mother serve as a model for interactions with close others. When interviewed about attachment and friendship issues, the child s quality of discourse emerged as the second most important influence on the quality of partnership representation. The interviews were conducted with the children at the ages of six, ten, and sixteen. They were analyzed for their coherency, clarity, and openness with which a child described and discussed his/her emotions and motives during social interactions with family and friends, as well as for his/her attachment orientation when presenting solutions for their problems in distressing situations or when having negative feelings. Partnership representation at age 22 had as significant predictors each of the assessments of discourse quality starting from the age of six years. In turn, discourse quality in adolescence could be predicted by discourse quality in childhood as well as by maternal sensitivity during childhood. The quality of discursive language when talking about close relationships seems to play a major role in the development of IWMs of attachment. We also applied the traditional assessments of quality of attachment to the children. In infancy, quality of attachment to the mother and father were assessed with the Strange Situation procedure: six-year-olds were given a projective picture test the separation anxiety test to assess their security of attachment representation; at adolescence, the Adult Attachment Interview was used. Compared to maternal and paternal sensitivity and support as well as, somewhat later, the child s quality of discourse about attachment issues, traditional attachment measures yielded fewer and much weaker relations to later partnership representation in this longitudinal study as well as in our second longitudinal study presented elsewhere (Grossmann, K. et al. 2005). The third major influence on partnership representation came from the child s experiences with his/her father. When the children were two and six

16 Universality of Human Social Attachment 213 years of age, fathers were observed in playful interactions with the child and rated on a scale for their Sensitive and Challenging Interactive Play (SCIP Scale; Grossmann, K. et al. 2002). This rating of the quality of their interactions was a significant predictor of later security of attachment representation at the ages of 10, 16, and 22 years, and it was a significant predictor of quality of later discourse about friendship representation at the age of 16 years and partnership representation at the age of 22 years. Later assessments of paternal sensitivity during play and paternal support, acceptance and empathic understanding of the child as reported by each side at age 10 also contributed significantly to later partnership representation (Grossmann, K. et al. 2005). Figure 10.2 presents the major results of the analyses of early influences on partnership representation at age 22. Maternal and paternal sensitivity and appropriate challenges while interacting with the preschool child were the two most important predictors. Traditional measures of security of attachment contributed very little, if anything at all. The quality of child discourse about attachment and friendship is not part of the graph because it was closely related to earlier maternal and paternal sensitivity. A wider view of attachment, which includes a child s experiences with both parents as havens of safety and secure bases in the domain of secure and playful exploration, was advocated quite early in the formulation of attachment theory (Bowlby 1982). Important for the development of a secure IWMof attachment is not only the extent to which a child s parents provide a secure base, but also the extent to which they encourage the child to explore from it. We fully endorse this wider view of attachment as it may yield a more comprehensive picture of the roots of a person s IWMof attachment. In modern sociobiological terms, these results indicate that quality of parental investment during the preschool years strongly affects a child s mental state with respect to the value of a close and supportive partnership many years later. Fathers as Trusted Companions during Exploration For a long time, attachment research has neglected the role of fathers as attachment figures and the child s negative emotions during exploration (e.g., wariness of mild fear). One reason for this may be that fathers challenge their young children s competencies, leaving the task of providing tender loving physical care and contact comfort to the mother (Parke 1996). This appeared true for the group of fathers that we studied in Germany: If a father wanted to be with his infant or toddler, they usually spent time exploring and playing. A supportive father who was neither overly demanding nor dominated their joint play seemed to provide the exploring young child with sufficient security through his supportive presence, while at the same time challenging the child s competencies. This appeared to motivate the toddler to master the challenges without anxiety. Positive, sensitive, and appropriate playful interactions often looked like vivid

17 214 K. E. Grossmann and K. Grossmann (a) 22 Y Security of Partnership Representation Quality of Discourse about Partnership p.05 p.01 16Y 1&2&6&10 &16 10Y 1&2&6&10 6Y 1&2&6 2Y 1&2 Maternal Sensitivity and Valuing Attachment (b) 22 Y Security of Partnership Representation Quality of Discourse about Partnership 16Y 10Y Attachm. Rep. p.05 p.1 0 6Y 2Y Security - SAT 1Y Attachment Quality Mother Attachment Quality Father Quality of Child Attachment and Representation Figure 10.2 (a, b) Relationships between maternal sensitivity and valuing of attachment and children s later partnership representation. SAT = security anxiety test. examples of a father knowing about Vygotsky zone of proximal development as the best way to foster his child s cognitive development. Therefore, when meeting difficulties, the children with their father at their side did not need to interrupt their concentration during exploratory and constructive play. The supportive, sensitively challenging companionship of fathers was found to have long-lasting effects on the child s socio-emotional and attachment development. Our findings suggest that infants develop attachment relationships to their mothers and fathers through distinctive pathways.

18 Universality of Human Social Attachment 215 (c) 22 Y Security of Partnership Representation Quality of Discourse about Partnership p.1 0 p.05 p.01 16Y Flexibility of Response to Social Rejection 10Y Clarity of Motives Perspectives for Action 6Y Clarity of Motives Perspectives for Action Quality of Discourse about Attachment Issues (d) Common 7% Unique to Mother 19% Unique to Father 7% Unexplained Variance 67% Figure 10.2 (c) Relationships between quality of discourse at age 6, 10, and 16 years and children s later partnership representation. (d) Explained variance in predicting partnership representation at 22 years: influences from mothers, fathers, and both parents (common). In Figures 10.3 and 10.4, the two kinds of security are contrasted for the purpose of demonstration in two pictures drawn by our team member, Sue Kellinghaus (Grossmann, K. and Grossmann, K.E. 2004). Figure 10.3 depicts the arrangement of the Strange Situation and may characterize paradigmatically a central feature of the young child s attachment to mother, as described by Ainsworth. A young child, if alarmed, will go to her mother for reassurance and relaxation. This function of the mother is usually noted as a haven of safety.

19 216 K. E. Grossmann and K. Grossmann Figure 10.3 Characterization of the Strange Situation, as described by Ainsworth (by S. Kellinghaus). However, by seeking proximity to the haven of safety, the child has to interrupt her exploration. If consolation is effective, the child will move off again to explore, demonstrating secure-base behavior. The Strange Situation procedure may certainly be a valid indicator of quality of infant attachment to mother but this method of assessment did not predict quality of attachment, friendship, and partnership representation beyond childhood in our two longitudinal studies. The second year of life may have been too early for the complex mental attachment organization assessed at 16 and 22 years (Grossmann, K. et al. 2005). Figure 10.4 depicts how psychological security is acquired during exploration through having a trusted and protective companion nearby. In this image, the father monitors the child s intentions and supports her actions at the site of exploration or Forschung. This constellation is typical of fathers interactions with their infants and children. In our longitudinal study, it turned out to be a valid and predictive indicator of the quality of the child father relationship and the development in the child of later attachment (Grossmann, K. et al. 2005). Both kinds of experiences with mother and with father were found to foster psychological security: a free, nonanxious access to others in case help, Figure 10.4 Characterization of the acquisition of psychological security during exploration in the presence of a trusted companion (the father) nearby (by S. Kellinghaus).