Clinical Implications of Attachment

Jay Belsky, Teresa M. Nezworski, Jay Belsky, Teresa M. Nezworski

• Publisher: Routledge
• Year: 2015
• ISBN:
  9781317838128
• Edition: 1
• Topic: Psychology
• Subtopic: Neuropsychology
• Index: Psychology

First published in 1987. This study records findings of a study group set up to explore a variety of issues related to attachment, including the predictive utility of Strange Situation assessments, the conditions under which insecurity is related to subsequent difficulties, the origins of individual differences in attachment security, and intervention strategies that might prove useful in ameliorating the developmental risks that appeared to be associated with insecure attachment relationships

I

GENERAL ISSUES

1

Clinical Implications of Attachment

Jay Belsky

The Pennsylvania State University

Teresa Nezworski

University of California, Santa Barbara

Only a decade ago, students of human development were routinely taught that individual differences in infant functioning in the first year of life were not predictive of later development. Despite this, many continued to pursue the question of the origins of individual differences that emerge after the period of infancy and that are so evident in human functioning across the lifespan. In fact, even in the face of all too consistent evidence indicating that individual differences were not stable from infancy to the childhood years, either in domains of intelligence or social and affective functioning, the collection of data that could radically alter this state of knowledge continued unabated.

In the absence of much evidence that individual differences were stable, clearly articulated positions emerged as to why such continuity should not be expected and could not be found. One point of view with respect to cognitive development was that the transition from sensorimotor intelligence, characteristic of the infant years, to symbolic intelligence, characteristic of the child, adolescent, and adulthood years, was qualitative in nature and, therefore, relatively impervious to stable individual differences (McCall, 1979). Additional theoretical support for this conclusion came from evolutionary arguments that asserted that development in the opening year or two of life was highly canalized and that unless the individual encountered gross deprivation in his/her rearing environment, normative variation in sensorimotor intelligence would not be predictive of variation in later intelligence (Scarr-Salapatek, 1976).

In the realm of social and affective development, stable individual differences were also strikingly difficult to document from infancy through preschool years. This led some to argue that the period of infancy was one of getting started and that it made little sense from an evolutionary perspective for variation beyond the range of the most serious dysfunction to have important implications for future development. In further developing this line of reasoning, Kagan (Kagan, Kearsley, & Zelazo, 1978) went so far as to argue, in his most eloquent and compelling manner, that the very notion of connectivity from early to later development was a figment of Western imagination grounded more in our cultural heritage and collective belief system than in the empirical world.

In view of this history, it seems reasonable to argue that over the past decade a virtual revolution has taken place in our understanding of early human development. This is because in both the cognitive and socioemotional realms of development there now exist sizeable bodies of evidence demonstrating that individual differences measured within the first year of life are, in fact, predictive of later development. With respect to cognitive development, a number of studies conducted in a variety of laboratories across the country have revealed that individual differences in information processing capabilities in the third through sixth months of life (and possibly earlier) are predictive—to a surprisingly large degree—of variation in psychometric intelligence (i.e., IQ) measured as late as the eighth year of life. Thus, infants who more quickly discriminate between novel and familiar stimuli, as has been revealed by a number of attention-habituation paradigms, score more highly on tests of intelligence as they grow older (see Bornstein & Sigman, 1986, for a review of the relevant research). In fact, these associations between early information processing and later intelligence are so robust and have been replicated so often that efforts are underway to develop and deploy screening tests to identify infants who might be at risk for later cognitive deficits that currently go unnoticed (Fagan, 1984).

In the realm of socioemotional development, it is the measurement of security of infant–mother attachment that has proved equally successful in documenting systematic associations between development at the end of the first year of life and subsequent functioning as late as the early school-age years. Studies carried out by Sroufe and his colleagues at the University of Minnesota (building upon the seminal work of Mary Ainsworth and the theorizing of John Bowlby), as well as investigations carried out in a number of laboratories around the county, clearly have demonstrated that infants whose relationships with their mothers can be characterized as secure as opposed to insecure using the Strange Situation paradigm generally look more competent as toddlers, preschoolers, and even as children beginning public school (see Bretherton, 1985 and Lamb, Thompson, Gardner, Charnov, & Estes, 1984, for reviews of the relevant data). All this is not to imply that individual differences in the security of attachment are deterministic of later development, but rather that theoretically meaningful associations between development in infancy and later functioning have been repeatedly established in studies focusing upon socio-affective functioning.

In recent years, the meaning of these associations has been the subject of great debate. Although they were anticipated on the basis of attachment theory, which stipulated that feelings of security and control growing out of the infant–mother relationship would contribute to the regulation of affect, the establishment of other social relationships, and the child’s negotiation of subsequent developmental tasks, it was never assumed that the child would be impervious to subsequent experience or that early attachment security would determine—in any unmediated fashion—the course of later development. Nevertheless, this is exactly what some critics of the theory and the research tradition it generated so often imply (Lamb et al., 1984), if not state directly (Skolnick, 1986). But as Sroufe makes eminently clear in his contribution to this volume, in which he clarifies many misconceptions that have been attributed to attachment theory, the basic assumption guiding attachment research is not that the relationship between mother and infant inevitably affects later development, but rather that the child’s initial relationship experience with mother probabilistically predicts later social development because it affects his/her expectations about others and relationships, feelings about self, and social skills used in other social contexts.

In reading contemporary criticism of attachment theory, one gets the sense not only that simplistic notions of development are being attributed to students of attachment, but also that there is little appreciation for the empirical contributions that this tradition has generated. The very fact that there is today a need to consider and debate why it is that early relationships predict subsequent functioning is in large measure a direct consequence of the success of the attachment paradigm in demonstrating that lawful relations do exist between security of infant–mother attachment and development in the toddler, preschool, and school-age years. That is, only because it has been demonstrated that there is continuity in socioemotional development has attention shifted to the conditions of such continuity. In our minds, the work of Erickson, Sroufe, and Egeland, (1985) is most instructive in this regard not only because it demonstrates an appreciation among students of attachment theory for these conditions of continuity, but because it indicates that the predictive power of assessments of attachment security, in large measure, depend on experiences that ensue in the post-infancy years. In showing that the association between insecurity and subsequent social difficulties is mitigated when mothers become more available and supportive of their offspring and that the association between security and subsequent competence is attenuated when the quality of maternal care deteriorates during the toddler and early preschool years, the work by the Minnesota group goes a long way toward modeling basic principles of attachment theory as well as of the developmental process more generally. More specifically, it underscores the fact that there is continuity in development when future experiences maintain developmental trajectories that have been established and that discontinuity characterizes the developmental process when experiences inconsistent with these trajectories are encountered.

This interpretation of the Erickson et al. research and this analysis of what is meant by the phrase “conditions of continuity” should not be read to imply that earlier experiences are insignificant and that only concurrent experiences are developmentally influential. The reason for this, clarified by the attachment theory, is that the child is by no means a passive recipient of experience (see Sroufe, this volume). Rather, the child is an active constructor of reality who both creates experiences and differentially attends to diverse information in his/her social world. After all, it was Bowlby’s contention in articulating the concept of “internal working models” that experience in the attachment relationship generates expectations of self and of others and that these are used to guide behavior and interpret experience. Evidence that this process is indeed at work in social development is suggested quite strongly by Dodge’s (Dodge & Richard, 1985) recent research on aggressive children. In demonstrating that such school-age boys are biased in their interpretations of the behavior of agemates by perceiving malicious intent and hostility when no such motives are present, it becomes clear that attributional processes not unlike those subsumed by the term internal working model do affect social functioning in important ways. Moreover, the fact that aggressive children are known to come from homes in which discipline is harsh and punitive (Parke & Slaby, 1983) is quite consistent with the notion that these models that guide behavior may well derive from earlier experiences in the parent–child relationship.

The possibility that there is a relationship between the processes that Dodge is studying and those that have traditionally been the focus of attachment researchers is suggested by the very research that stimulated the convening of a study group sponsored by the Society for Research on Child Development concerning Clinical Implications of Attachment, which resulted in the current volume. As a result of two separate investigations documenting a link between early insecurity and subsequent behavioral problems (Erickson et al., 1985; Lewis, Feiring, McGuffog, & Jaskir, 1984), it became clear that the concept of attachment, which had proven so useful to developmental psychologists studying normal developmental processes, might well have some clinical implications. That this might indeed be the case was certainly evident in Bowlby’s early writings, but it took the research of Erickson et al. and Lewis et al. to convincingly demonstrate a linkage between early insecurity, as measured in the Strange Situation, and subsequent behavioral problems. In neither of these studies, it is important to note, was it the case that insecurity inevitably led to externalizing or internalizing behavioral problems; the seminal contribution of this work, however, was the empirical demonstration that childhood problems long thought to have their roots in difficulties in the early infant–mother relationship were indeed related to insecurity of attachment.

Actually, when one carefully examines earlier research on the developmental correlates of attachment security, it becomes apparent that the associations documented by Erickson et al. and Lewis et al. were very much foreshadowed by earlier findings in the literature, though never cast in terms of behavioral problems per se. Consider, in this regard, the fact that Matas, Arend, and Sroufe (1978) found, in studying toddlers’ behaviors during a problem-solving task, that those with histories of anxious-resistant attachment were whiney, negativistic, and easily frustrated, and that those with anxious-avoidant histories were neither compliant nor cooperative. Similarly, Maslin and Bates (1982) observed that 2-year-old who had been classified as insecurely attached to their mothers at 13 months of age engaged in more conflict with their mothers than infants evaluated as securely attached.

When one considers these findings linking early insecurity with behavioral functioning that is often associated with referrals to child guidance clinics, and then reflects upon the recent emergence of interest in developmental psychopathology, it is not difficult to realize why we believe that one of the major contributions of research and theory on attachment has been the drawing together of students of normal and abnormal development. In fact, we wonder whether the resurgence of interest in developmental psychopathology that we have witnessed in the past few years (Cicchetti, 1984; Sroufe & Rutter, 1984) would have taken place without the seminal contribution of attachment research. This is not to say that the basic concepts underlying the notion of developmental psychopathology have not been around for some time, but only that it took evidence documenting associations between normal and abnormal processes, which attachment research did much to provide, for the interrelation of the normal and dysfunctional and the developmental roots of psychopathology to capture once again the attention of mainstream developmental psychology.

In the chapters contributed to this volume, clinical implications of attachment theory are examined. In overviewing some of the primary contributions of these chapters we find it useful to think in terms of characteristics, consequences, and determinants of attachment security, as well as to consider the direct clinical applications of concepts derived from attachment theory in the provision of services to children and families. In the remainder of this chapter we outline what we view as some of the most important issues addressed in these chapters.

CHARACTERIZING VARIATION IN ATTACHMENT SECURITY

For many years now it has been traditional to employ the standard Strange Situation laboratory procedure when infants are 12 to 18 months of age in order to assess the security of the infant–parent attachment relationship. In distinguishing secure and insecure relationships, there has been some confusion about the interpretation of maladaptation applied to insecure attachments. As Sroufe, Crittenden, Lieberman and Greenspan, and Bates and Bayles all make clear in their chapters, the term maladaptation is employed within the attachment framework in terms of the child’s future development. All of these authors recognize the fact that insecurely attached infants have established relationships that must be considered adapted to the circumstances of their rearing, even if they prove to be problematical as they move into the world beyond the family. Thus, insecure relationships are considered to be functional in that they serve to protect the child against anxiety, which arises in the face of a caregiver who may be less than optimally available. When seen in this light, avoidance, for example, serves as a strategy for avoiding anger that may evoke negative responses from the caregiver (see Cassidy & Kobak, this volume).

Until recently, behavior observed in the Strange Situation was used to characterize the infant–mother relationship in terms of one of three kinds of attachments: secure, anxious-avoidant, and anxious-resistant. In recent years, investigators working in a number of labs have identified subgroups of infants that fall outside of these traditional groupings. One particular subgroup that has attracted a great deal of attention is comprised of infants displaying both heightened avoidance of the mother when reunited with her following a brief separation as well as heightened resistance. As Crittenden demonstrates in her chapter, unless an A/C coding category is employed to capture such relationships, these children can be classified erroneously as secure. Particularly noteworthy in this regard is Crittenden’s findings that such infants are disproportionately likely to have been abused and/or neglected. Such findings merit special attention from the standpoint of clinical concerns particularly when it is noted that children of depressed mothers also are disproportionately likely to be classified into this category when it is available for consideration (Radke-Yarrow, Cummings, Kucynski, & Chapman, 1985). The possibility that this classification is especially likely to be associated with disturbances on the part of the mother is strongly suggested by Speiker and Booth in their chapter, which devotes exclusive attention to the caregivers of infants with A/C classifications (see the following).

ORIGINS OF ATTACHMENT SECURITY

This discussion of the maternal correlates of attachment relationships that are classified A/C draws attention to one of the major debates that has always surrounded consideration of individual differences in infant–mother attachment security; namely, what are the determinants of variation in the quality of th...