Developmental Theories of Parental Contributors to Antisocial Behavior
Introduction
As noted by Loeber and LeBlanc (1990), most theories in criminology do not take a developmental perspective. Even when interested in change, the tendency is to study processes going on at each of two times rather than within subject change between the two time periods, a more developmental approach. Knowledge of the underlying developmental process contributes to the study of etiology in antisocial behavior, and also to the ultimate goal of prevention. Pointing out intervention efforts that have had adverse effects on antisocial children, Bell (1986) argued that the most defensible basis for intervention is a well understood delineation of the developmental pathway to the disorder, including changes in family and other contexts that affect the course.
Given the argument for a developmental perspective, the first question is which of the many substantive areas associated with antisocial behavior are sufficiently promising that they should be given the benefit of a developmental approach. Two reviews (Loeber & Dishion, 1983; Loeber & Stouthamer-Loeber, 1986) have come to the conclusion that harsh, inconsistent discipline, inadequate supervision, parental rejection, and lack of involvement with the child have shown the most power in concurrent prediction during childhood and the most consistency across a broad range of studies. The findings concerning parental rejection and lack of parental involvement can be subsumed under the broad category of unresponsiveness and lack of sensitivity to the child's needs (Egeland & Farber, 1984; Lewis, Feiring, McGuffog, & Jaskir, 1984). Thus family factors involving discipline practices and the quality of the parent-child relationship are at the forefront of results from meta-analyses.
The next question arising from a developmental orientation is what processes led up to the difficulties in the quality of the parent-child relationship, to harsh, inconsistent discipline, and to a lack of parental involvement? In the case of disciplinary practices, Bell and Chapman (1986) concluded that experimental manipulation of child behavior can produce variations in parental power assertion in artificial parent-child pairs. Thus it is reasonable to speculate that when faced with an overly assertive, goal-directed child, a parent who lacks firmness and adequate discipline techniques might vacillate between yielding and power assertion such as demanding compliance, then, occasionally out of desperation, resort to harsh discipline. Further, Lepper (1982) has reviewed research demonstrating that failure to get minimal compliance increases the difficulty of obtaining compliance in subsequent interactions with a child. From the formulations of Reid and Patterson (1991), such a parent-child relationship might develop into coercive cycles of interaction, depending on the techniques parents use.
From the foregoing, it is no longer defensible to assume that the influence of the child is so negligible that it is only necessary to work out the effects of the parent. Family therapists have long since recognized the need to work with the complex reciprocal influences within a family (Ackerman, 1954; Jackson, 1957), but investigators of normal socialization within developmental psychology have only more recently, for methodological reasons, identified the flaws in models of unidirectional influence (Bell & Harper, 1977). Now most investigators would challenge the assumption that it is more reasonable to interpret correlations between parent and child behavior as indicating the effects of parents rather the effects of children. Furthermore, when evidence from behavior genetics and experimental studies was included along with the usual correlational studies in the most recent developmentally-oriented review of child and family contributors to antisocial behavior (Lytton, 1990), the author concluded that not only do the effects of the child's characteristics have to be considered, but that they have a stronger influence than those of the parent.
A theory must also be transactional (Sameroff, 1990), meaning that out of the continual interplay of parent and child, behavioral outcomes emerge that are not attributable to either participant alone. The most flexible theories will also be transformational or epigenetic, meaning that developmental changes arising from other sources than interaction may also take on forms not shown in previous stages.
From a developmental perspective it seems almost redundant to stipulate that a theory which only applies to one period in the life of the antisocial child, or one that postulates a single factor that is assumed to have persisted throughout the pathogenic process, is no longer likely to be viable. Even though stimulating and heuristic, a theory must move beyond the assumption of simple continuity. Change is the hallmark of development, constant and unrelenting, from the infant who is working on the attachment process, to the toddler who is engaged in differentiation and development of a self concept, to the preschooler learning to function in a peer group. There are nearly continuous universal maturational changes in the individual child, as well as maturation specific to an ethnic group or to family lines.
In addition to processes going on in the family, changes are entering the parent-child system from outside the home so that the infant or child has to meet continuous demands for improvement and control of behavior. Developmental psychology is increasingly turning toward new and complex theories that will accommodate these transformations.
Dynamic systems theory (Thelen, Kelso, & Fogel, 1987) is an important model that has emerged in the last few years out of chaos theory in physics, organismic psychology, Piagetian theory, and general systems theory. This model challenges linear growth models, such as social learning theory, and successfully predicts discontinuities in the physical growth process. Developmental psychology is only one of several disciplines actively pursuing this theory and a national society has already been formed. From this theory, we would expect complex changes in the development of an antisocial child rather than the mere decrease or increase in isolated behaviors that are reinforced or modeled to elicit imitation. One other model based on general systems theory has already been applied to the conceptualization of risk for schizophrenia and other emotional disorders (Sameroff, Seifer & Zax, 1982). Within the field of developmental psychopathology Sameroff''s papers on the transactional approach are very frequently cited.
Six existing theories of parent-child interaction or family process will be reviewed to determine whether: (1) the age range is adequately covered, (2) the focus is on microanalytic or general behavior, (3) a reciprocal or transactional approach to parent-child interaction is adopted, and (4) the theory accommodates normative or maturational changes in the child, as well as transformational change (sometimes referred to as heterotypic or epigenetic change). Items 1 and 4 derive from the reasons already given in this paper for the importance placed on a developmental approach, while 3 and 4 represent the value placed on conceptual approaches that meet the challenge of complex growth processes, and are currently commanding considerable interest amongst social and behavioral science theorists. Item 2 is included since it taps a basic feature of all theories, the units of analysis used. While holding out the foregoing desiderata for conceptualizations in this area of research, it is nonetheless appreciated that the field of research on antisocial behavior, just emerging from decades of empirical work, is not likely to have produced complete, formal theories. Table l provides a list of six theories, of which two involve social bonding, two involve attachment, and two involve social learning theory. After these are reviewed, we propose a new bridging model to link two theories that come close to overlapping in developmental periods covered, and that provide opportunities for integration of concepts.
The review does not cover the extent to which each theory incorporates the effects of social context factors on the development of antisocial behavior, nor transitions in the child's life created by the broader society beyond the family, since the review concentrates on theories of parental behavior. Also, the theories could not be differentiated readily on the basis of how well they accommodate context. A variety of context elements are represented in these theories, each equally justifiable from the limited perspective of this review.
Social Bonding Theories
Hirschi
Glueck and Glueck (1934) were some of the earliest to detect hostility or lack of warmth in fathers, and hostility or indifference in mothers of delinquent children, but Hirschi (1969) was one of the first to integrate these and other related findings into a theory of how parents contribute to antisocial behavior. Hirschi's theory is a variant of classical control theory in sociology that places emphasis on social sanctions and external control. Hirschi believes that lack of self-control is the key element and that this is due to the failure to form an attachment to parents in early development, which leads, in turn, to failure to form an attachment to other conventional individuals, failure to make a commitment to conventional lines of action and conventional activities, and lack of belief in the legitimacy of the moral order. However, in this theory attachment and bond are often used interchangeably to refer to the larger society beyond the family. The terms are synonymous with "subscribe to," "believe in," or "act in accordance with," whereas in the developmental literature the term attachment is primarily applied to the various social behaviors of an infant that have proximity to a mother or primary caregiver as a predictable outcome and which reach a peak between 12 and 18 months. In this review the term attachment will be reserved for the process in early infancy, while bonding will be used to refer to "attachment to society" since most readers are accustomed to the latter term used in control theory and the mother's bond to her infant does not enter into the review.
In the early formulations of the theory (Hirschi, 1969), the attachment of the young child to its caregiver was considered the key to socialization. However, the most recent statement of the role of parents makes very little mention of the child's attachment to its parents (Gottfredson & Hirschi, 1990). Instead, the theory concentrates on characteristics of the child that presumably result from its bonding. The current theory attributes lack of control to lack of parental concern for the welfare and behavior of the child (as manifest in hostility toward the child and lack of warmth). Gottfredson and Hirschi say that parental concern about trouble in school and poor school performance must also lead to supervision, parental insistence on meeting goals of socialization, which, in turn, is thought to move the child psychologically from external, monitored control to internal control. How the latter process leads to internal control is not spelled out.
From Table l it can be seen that the developmental period or phases to which the theory is to be applied extends from early infancy to late adolescence, though the authors own investigations primarily involved high school youths. The theory is essentially put together from findings in the many studies already in the literature (Glueck & Glueck, 1934; McCord, McCord, & Zola, 1969; Robins, 1966; West, 1969), and these cover a wide age range, but are primarily concentrated on early childhood and adolescence. Consistent with this literature, parent behavior is seen in the form of traits.
Hirschi's theory is not reciprocal. The postulated parent effects are unidirectional, and the theory contains no provision for child effects on parents. Nor is the theory transactional because the outcome of the family process is a characteristic seen as residing in the child, such as low self-control and risk taking. Sameroff and Chandler (1975) would describe the child's risk taking as a product of the child in its family and school contexts, and might point out that in the company of deviant peers the child may show less risk taking and show more self-control.
In its original form (Hirschi, 1969), the theory postulates a change from a child lacking a secure attachment to its parents to a somewhat older child who has a weak bond to society. A developmental progression involving different forms of behavior is implied and thus, to this extent at least the theory is transformational. In fact at earlier ages it appears that a factor resembling maternal responsiveness might be implicated in the origins of antisocial behavior, as lack of parental concern or warmth results in lack of self-control on the part of the child. Later on during middle childhood and adolescence, parental management techniques are conceived as more important. During this latter period, lack of supervision and failure to monitor seem likely to condone the antisocial behavior in the child's mind, and prevent the growth of internalizing control. However, the theory does not tie its concepts into normative changes in the child other than attachment.
Hawkin's Social Developmental Model
One other social bonding theory has been devised to explain the etiology of substance abuse. It is proposed as a basis for intervention but is also considered relevant to antisocial behavior because of the close link between the two disorders. Hawkins, Lishner, Catalano, and Howard (1986) emphasize the importance of the child's bond to society and use concepts from social learning theory to spell out the nature of the process that presumably underlies formation of a bond. The theory also postulates a developmental progression of bonding to family, school, and peers in the period from early childhood to late adolescence. The social learning components needed in family, school, and peer relations are availability of opportunities for prosocial activities, involvement in these activities, and the existence of rewards for showing conventional social, cognitive, and behavior skills. The theory states that attachment, commitment, and belief in the family do not develop when parents are inconsistent and lack involvement with the child. To this extent, the theory makes some use of traits, but in the main the theory conceptualizes parent behavior as types of influence. Turning to the child, the result of the failure to provide positive influence is a child who fails to perform skillfully in conventional settings and lacks skills in avoiding unconventional settings. Bonding to the family is necessary but not sufficient for bonding to the school, where the 4 elements listed must again be present. Bonding to the school is also necessary but not sufficient for bonding to the non-deviant peer culture, in which the same elements must be present. In summary, for the purpose of the present paper, the theory indicates that parents influence their children by providing prosocial activities in the home, opportunities to motivate them to participate in these activities, and a pattern of rewards which ensures that they acquire conventional skills.
The theory goes beyond negative trait descriptions of parents of antisocial children, such as "inconsistent, "punitive," and "ineffective," and concentrates instead on what opportunities parents are providing and what behaviors are reinforced. However, the effects of a troublesome child on prosocial activities and rewards have not been considered. "Difficultness" or "irritability" may be the way in which a biological contributor acts through the child on the parent-child interaction. Inconsistent rewards and harsh punishment could well be a response of some parents who become exasperated when their management techniques fail to secure compliance or prosocial behavior. Antisocial children often show a paradoxical negative reaction to negative reinforcement.
If inconsistency and harsh punishment were a response to an increasingly angry and defiant child, rather than an original or earlier determinant, we would not expect to find relationships between the negative parental traits and later outcome. In fact, only one of four studies with relevant data (McCord, 1978) has obtained longitudinal relations, leading some reviewers (Maccoby & Martin, 1983, 42-44; Lytton, 1990) to conclude that the concurrent relation of these particular negative parental behaviors does not indicate a causal effect on the child.
A theory that is not reciprocal cannot be transactional; the social developmental model lacks components in both of these areas. However, as can be seen in Table l, the model is transformational in that it distinguishes very general periods of development, the last being centered around peers. This postulated sequential order is in question. The peer culture outside of the home exerts effects as early as day-care in infancy. Even if there has been no day-care experience, preschoolers would be starting their exposure to non-sibling peers at the time of their entry into kindergarten or first grade. The authors of the theory would undoubtedly accept this fact of early peer experience, but point to late adolescence as the period in which peers exert maximal influence toward deviance or conformity.
Although the theory is transformational and also identifies the very different contexts for learning from the home to school, it does not show any adjustment to normative changes in the child. In all areas of development from motor to cognitive function the average child shows many maturational changes between early childhood and late adolescence.
In summary, with respect to its theory of parental behavior, the social developmental model has moved beyond the original social bonding models and taken steps toward goals that other theories should also follow. There is minimal reliance on trait labeling of parents, and there is better specification of what parents should actually be doing with children to forestall antisocial behavior.
Attachment Theories
Minnesota Studies
Sroufe's (1983) conceptualization of Bowlby's (1969, 1973, & 1980) and Ainsworth's (Ainsworth, Blehar, Waters & Wall, 1978) attachment theories provided the groundwork for a series of studies on normal infants, then a follow-up on a high risk sample that yielded enough antisocial children by age 8 that an attachment theory of parental contributors could be tested. Since the theory reaches down into early infancy to identify precursors, and has been elaborated in many empirical and theoretical articles, it will be accorded much more space than other theories.
The core of Bowlby's nomothetic theory is that all infants attach to a mother or primary caregiver if this individual is regularly available to interact with them. The infant's initial role is to signal its need for care and, later, maintain the caregiver in proximity by its own locomotion. Differences in the quality of attachment, however, are considered to be due to the caregivers sensitivity and cooperative, non-intrusive care. According to Bowlby's (1973) concept of the "internal working model," infants derive from early interactions their basic expectations about the physical and interpersonal environment, including the availability and responsiveness of others and their effectance and worthiness. Infants who have experienced sensitive and dependable care (responsiveness to cues of its needs and condition) will not only expect that the caregiver will be the source of their safety, but also that they can bring the caregiver to them or achieve security by going to the caregiver. As a result, the infant will be secure in the attachment relationship. The same body of theory postulates that securely attached infants function more harmoniously with their mothers in compliance situations because these infants are motivated to please their mothers and attend to what she approves or disapproves (Stayton, Hogan, & Ainsworth, 1971). Further, the attachment provides a secure base from which the infant explores the environment, generating stimulation that promotes cognitive development, sharing of new experiences, and gaining the capability of cooperation with the mother in joint problem solving or work situations (Sroufe, 1983).
From study of the infant's behavior in a standard separation and reunion situation, attachment types have been identified that are associated with normal and deviant development, and assumed to result from variations in the mother's responsivity and sensitivity .
In Type "A" (anxious/avoidant) attachments, the infant avoids contact with the parent during a sequence of separations and reunions. Secure infants (Type "B") seek comfort and proximity, then increasingly return to play. In Type "C" attachments the infant shows anger and resistance when the parent returns, typically seeking contact, but is not readily comforted. Finally, a new category, the disorganized/disoriented Type "D" category has been created to describe the aberrant patterns demonstrated by maltreated infants. Some additional differentiating characteristics of the "A" pattern should be noted, since it is the type most frequently implicated in later externalizing behavior. Reviews by Main, Kaplan, and Cassidy (1985) and Sroufe (1983) include the following features: (1) active directing of attention away from the mother, (2) avoiding the mother in cooperative tasks, and (3) appearing overcontrolled. The latter is not surprising if we consider that expression of anger directed toward an indifferent caregiver might decrease even the minimal responsiveness being shown to the infant and thus increase feelings of vulnerability.
Very different models of the self, others, and relationships would result from inconsistent, haphazard care or rebuffs of the infant's efforts to make contact. The infant's internal working models function as guiding frameworks for further interaction with the environment. These models involve selecting and interpreting experience as well as how one relates to others. Later, if insecure-avoidant children expect to be rebuffed by their parents, they may avoid them when they should be coming to them, and thus the parents lose the kind of contact that is necessary to monitoring. Insecure-avoidant children may interpret neutral or even friendly behavior as hostile, and may show completely inappropriate aggressive behavior. Having constructed working models of their relations, they then bring these to bear on each new experience.
Egeland & Farber (1984) have reported that mothers of infants who shifted from secure to avoidant attachments between 12 and 18 months scored high before the infant's birth on self-reports of aggressiveness and defendence, but low scores on social desirability. A similar pattern was found for defendence and social desirability 3 months after birth. The investigators characterized these mothers of infants who shifted by the end of the 18-month period as irritable, tense women who lacked interest in their infants, handled them only when necessary, provided little opportunity for reciprocal interaction, and lacked skills in caregiving. This is a consistent and meaningful set of transformational relations from prenatal, or early postnatal maternal personality, through to the mother's feelings about the maternal role, and the nature of her caregiving.
Reports on long-term follow-ups made it possible to relate this indifferent and neglecting pattern of maternal care to the child's later noncompliance and externalizing behavior. In the only study follow-up using a low SES sample (Erickson, Sroufe & Egeland, 1985), avoidant attachments in particular were associated with later noncompliance and hostile acting out behavior. This association has been maintained in a follow-up at age 7 to 8 years (Renken, Egeland, Marvinney, Mangelsdorf, & Sroufe, 1990), but only for boys who were rated type "A" at 18 months of age. On the other hand, Fagot and Kavanagh (1990) found no relation between avoidant attachment classification at 18 months and child externalizing behavior at 24-48 months. This study also provides evidence to support the contention of Bates, Maslin, and Frankel (1985) that the greater number of protective factors present in middle-class homes may avert the emergence or severity of adverse outcomes from insecure attachments. Thus it is possible that the more severe problems of compliance often found in low SES samples are associated with "A" attachments, whereas the less severe problems of compliance in middle-class samples are associated with both "A" and "C" attachments.
The relations between prenatal or neonatal measures and later attachment outcome in the studies by Egeland and Farber (1984) have been dismissed by one review article (Lamb, Thompson, Gardner, Charnov, & Esler, 1984) on the basis that most longitudinal correlations were shown to the 12-month or 18-month assessments of attachment, but not to both. This review overlooked the fact that a major change in the child's physical, perceptual, and cognitive functioning occurs with the onset of walking during this period. Transformations, not literal continuities, should be expected of longitudinal data involving such an important change. The problem of methods in this study that should be a basis for concern involves the differences between samples and how attachment was measured. For example, the central hypotheses comes from the study by Egeland and Farber (1984) in which prenatal, 3- and 6-month correlates of later 12- and 18-month attachments were studied, as well as shifts in quality of attachment between these ages. On the other hand, the relation between avoidant attachments and preschool noncompliance came from a study of infants that showed no change in attachment (Erickson et al., 1985). The study that reported relations between early avoidant attachment and aggressive antisocial behavior at age 7 to 8 years came from a study in which later correlates of both 12- and 18-month attachments were analyzed, but only the 18-month data proved to be predictive. Overall, the hypotheses were confirmed and an exciting set of findings emerged, but it must be kept in mind that the attachment data were treated differently in the different studies, and the data come from one report on a middle-class sample and 3 reports on different segments of a follow-up carried out with a high-risk low-income sample.
It should be no surprise to clinicians to encounter negative consequences of neglecting mothers, but to know that these mothers were different even before the birth of their infants is new. It is also new that one specific effect of these mothers is on the kind of attachment their infants form with them. This tracing of a pathway up to age 7 to 8 years in the follow-up studies touches the lower bound of the period from 7 to 9 years in which Patterson and colleagues (Patterson, Debaryshe, & Ramsey, 1989) speculate that the coercive cycles leading to antisocial behavior have their origins.
We have already mentioned the wide age range intended to be covered by the conceptualizations of the Minnesota investigators, and the next feature noted in Table l is the extent to which the parental behavior is specified. The description of lack of responsiveness and lack of sensitivity in mothers of avoidant infants is quite detailed, though it does not reach the level of minute to minute interaction.
The attachment theory as articulated by Sroufe (1983) is neither reciprocal or transactional. This extension of attachment theory overlooks Bowlby's (1969) original conceptualization of the prototypical infant in the average expectable environment, who signals to bring the caregiver. Rather, Sroufe turns to a model of individual differences, accords no causal role to variations in infant behavior such as those produced by temperament, while attributing the primary role to that of the sensitive and responsive mother. The roles of mother and infant never meet in Sroufe's theory. A secure attachment is something created by a mother, not a result of a reciprocal or transactional process involving both mother and infant.
In contrast with its unidirectional conceptualization of the mother-infant interaction, the theory as articulated by Sroufe, transformational in the sense that the infant builds a model of relationships out of its interaction with caregivers, and in that the effects of the parental behavior are seen in the emergence of new behavior, rather than in literal continuities. Furthermore, developmental changes in the infant are clearly identified in designing testing situations at different ages.
Greenberg's Cognitive/Affective Model
Greenberg and colleagues (Greenberg & Speltz, 1988; Marvin & Greenberg, 1982) have advanced an explanation of antisocial behavior that extends the attachment model from early infancy through the preschool years. It is an intensive developmental model that also provides very specific examples of the nature of parent-child interaction during the preschool years. The theory is shaped around 9 different classes of developmental change that occur in the period from 2 to 4 years, how the parent and child interaction is altered as a consequence of these changes, and how changes in the child's representation of its world are carried through into each of these stages. The basic developmental processes that occur in most infants are elaborated first, then the precursors of antisocial behavior are constructed against this background of normal behavior.
According to Greenberg's view of normal development, the secure infant whose mother was sensitive and responsive during the period of early infancy, manifests her interest and involvement in a very different way during the early preschool period. She (or other primary caregiver) is expected to be engaged in building what is termed an emergent partnership by talking for both herself and the child about joint activities, interpreting the activities, and labeling the child's and her own emotional states, thus scaffolding the child's cognitive and emotional abilities to levels that would not be attained with only simple caregiving. The mother is also being flexible in setting limits during this period in which the child literally assaults the parents and house with its newly found mobility and need for autonomy, but she is firm in enforcing the limits once set.
Toward the end of the preschool period the secure dyad should have attained a "goal corrected partnership" in which it is possible to negotiate goals and make plans. The mother and child can work together effectively in household work and problem situations as the child has become more tolerant of its own emotions, and shows more self-control as a result of the continuous communication with the mother about anger and other emotions. The secure child should have constructed a stable internal representation of self in relation to the interpersonal context in which s/he is embedded, a representation that Bowlby termed a "working model." This representation is carried forward into the next developmental stage that involves teachers and peers. The child is thought to be constantly engaged in revising and expanding this working model throughout the preschool period in which there are rapid changes in locomotion, language, cognition, impulsivity, self concept, autonomy, and relatedness.
In the main, hypotheses about parent functioning in antisocial behavior have not been tested empirically but warrant attention because they are built on a very solid integration of the literature on attachment and normal development. The advocates of this model contend that disruptive behavior in preschoolers is an attempt to get attention or control the behavior of unresponsive or unpredictable caregivers who have not built up good communication relative to goals and plans. These latter have to do with the minutiae of daily living that look minor to an adult but loom large in the child's view of the world. Examples are cleaning up toys, eating without spilling, or working out how brief or more extensive separations are going to be managed.
The failure to attain a goal-directed partnership is also assumed to reveal itself in failing to share or in being intolerant of emotions. The insecure child feels a mixture of fear, anger, sadness, and frustration at the inability to get or maintain predictable, supportive contact, meaningful communication or cooperation. The child does not understand these feelings and cannot get help in managing them. Lacking tests of the theory, the authors can only speculate that laissez-faire or authoritarian patterns of parental behavior, as well as unpredictable variations between overcontrol or punitive parental behavior that "jerk" the child back and forth, could all lead to anger in the child and coercive cycles of interaction.
Comparing this cognitive and affective theory with others in Table l, it is apparent that it is unique in that it concentrates so intensively on the preschool period. It is also unique in that it takes each developmental change in the child into account, as well as major changes in functioning of the dyad such as progression from an emergent partnership to a goal-corrected partnership. Thus the theory is one of the most transformational of all those discussed, though limited to a specific developmental period. The theory is also reciprocal and transactional in that it accepts the two-way direction of effects in interaction, and sees the child as a carrier of effects from its temperament, context, and previous interactions. Some of the theories' hypothesized internal connections have been examined empirically (Marvin, 1977), but the overall theory has yet to be tested in longitudinal studies. It is a relatively new model that should yield testable hypotheses.
Social Learning Theories
Patterson
Patterson and colleagues (Patterson, DeBaryshe & Ramsey, 1989) distinguish between an early and late starter pathway to antisocial behavior. The key processes involved in the latter get underway during early adolescence. Both processes involve deficits in family management skills, but the early starter model begins with a troublesome child and/or parents who lack firmness and skills in discipline, leading to coercive cycles of interaction when the child is from 4 to 9 years of age. In the case of the troublesome child the cycles are presumed to originate in hyperactivity or in any other condition that produces irritability, to which the parent responds aversively but ineffectively, thus rewarding the child's own aggressive and aversive behavior. Thus a stable system of mutually coercive cycles of interaction is built up out of the countless daily interactions of early childhood. The goals of the child and parent differ at times so that conflict is present even in normal families. However, if daily coercive cycles of interaction predominate, the child learns how to be abrasive, and does not receive training in social skills, so that aggressive behavior and lack of social skills are carried into the school setting, where classmates soon become alienated. Peer rejection and neglect detracts from learning, even if the inability to comply with instructions in the classroom has not already impaired learning.
In middle childhood, the antisocial child spends more time unsupervised by the parents, who are irritable and ineffective in the home and thus unlikely to find out what is going on outside the home. The parents learn less about the child`s conflicts with peers and teachers, and thus are unable to monitor out of home activities (Reid & Patterson, 1991). The child's rejection by the mainstream peer culture and failure in school leads to contact with a deviant peer group that is almost always available, and with whom training in coercive cycles provides a basis for affiliation. The lack of communication with the parents then precludes guidance away from the deviant culture, and permits deepening commitment to the group.
Although Patterson's data gathering largely concerns a period much later in development than the period to which the theory is extrapolated, the explanatory system meets most criteria for a comprehensive developmental model. First of all, the behavior of parents is specified at the microanalytic level in the minute to minute interplay between the aggressive child and the irritable, ineffective parent. The theory is reciprocal and transactional even at the microanalytic level in that the child's aggressive behavior elicits an aversive parental response that further reinforces the aggressive child behavior (Patterson, DeBaryshe, & Ramsey, 1989). The theory is transactional even at the level of traits since the authors speculate that a troublesome child in the hands of a highly effective parent or a competent and easygoing child with a very ineffective parent might not emerge with antisocial behavior (Reid & Patterson, 1991). Thus at the level of perspectives the authors are flexible and ready to recognize that the direction of effects might flow from either the child or the parent at any given time and that the results of the parent-child interaction are a product of the action of both. However, both in the design and analysis of the data, it is apparent that this is primarily a social learning theory.
Patterson's theory is transformational in that different forms of deviant behavior occur at different points in the development of the disorder, but with one exception, the theory does not integrate normative changes in behavior that occur in most children with the changes expected in the antisocial child. The exception is the effect of the child's shift into pubescence. This shift is seen as stressing parents with marginal family-management skills.
One of several valuable contributions of Patterson's theory is that it suggests a mechanism by which a wide variety of correlates such as mental health problems in the home or hyperactivity in the child have their effect on the development of antisocial behavior. According to this theory, these variables owe their correlations to the extent to which they cause disruption in parenting practices. Considerable support for this simplification of the causal matrix is marshalled from the pattern of relationships between the many potentially relevant variables that have been analyzed (Patterson, Capaldi, & Bank, 1990).
Martin's Empirical Model
Unfortunately, investigations that have attempted to trace the early determinants of the coercive cycle are scarce. Only one study could be located that attempted to extend the concept of coercive cycles into earlier age periods than the school-age antisocial children on which it has been based. Using an intensity-matching model of interaction, Martin (1981) obtained unusually clear longitudinal relations based on brief laboratory measures, making it possible to trace the somewhat dissociated interactions of uninvolved, unresponsive mothers with their aversively demanding 10-month-old infants through to 22 and 42 months, when noncompliance and coercive cycles of interaction were identified. Two variables, lack of maternal responsiveness, infant demandingness and, importantly, their interaction term were identified at 10 months of age as precursors of coercive child behavior at 42 months. Martin's data fit Patterson's transactional model well, as dyads that evidenced both unresponsive mothers and demanding infants at 10 months were more likely to demonstrate high rates of noncompliant behavior at 22 months, and both variables were independently related to coercive child behavior at 42 months. Martin's "causal" paths indicate that, from the first year forward, coercive cycles emerged slowly out of the cumulative effects of the infant's demandingness and the mother's lack of involvement. Martin's data also bear some resemblance to findings from the Minnesota project. One of the two salient predictor variables was lack of maternal responsiveness, which is considered to be a primary determinant of security of attachment in several studies reviewed by Sroufe (1983).
Martin's conceptualizations are backed up by very concrete, specific behaviors directly observed in mother-child interaction, and their reciprocal, transactional nature are clearly evident in the findings reviewed above. The transformational nature of the findings is also clear in that the early mother-infant interaction started in one form (infant demandingness and lack of maternal responsiveness) and ended up in quite another (coercive cycles). Furthermore, a developmental rationale taking into consideration normative changes in all infants, was provided for each period of data gathering. For a conceptualization based on a single study, Martin's is uniquely informative because of its use of directly observed interaction data to identify specific behaviors, and analyses that permitted reciprocal, transactional, and transformational findings to emerge.
Unfortunately, security of attachment was not measured in the Martin study, so the voluminous literature from attachment and antisocial children cannot be aligned empirically. Nor can Martin's sample be considered at risk for adverse development. They were primarily from middle- to upper-middle class families with high educational attainment. Despite these reservations, just as in the case of Sroufe's and Egeland's low-income families, continuity was found between parenting variables, and effects were stronger for boys.
A Bridging Developmental Model
We propose a bridging theory that draws from previous research but weaves it together in order to fill the conceptual gaps that exist between research on prenatal maternal personality, temperament, and coercive parent-child interaction. The bridging theory is frankly speculative; however, the speculations are primarily an outgrowth of empirical work. The model is specifically designed to accommodate the mother-son interaction patterns of families from low socioeconomic levels. Our reasons for limiting the theory to males from low socioeconomic backgrounds are based on the higher base rates in both cases and the fact that previous longitudinal studies have reported quite different gender patterns over time. We see no sign in the data we have reviewed that a theory is likely to emerge that will encompass a similar developmental pathway for both sexes. In the case of socioeconomic levels, it is recognized that most low income families are able to provide economic and social supports for their children despite fewer financial resources and lower educational attainment. However, even though many low SES families meet these challenges successfully, children from such homes continue to be at greater risk than their middle- or high-SES peers for developing many types of child and adult psychopathology, including antisocial behavior. Further, the theory is only proposed as one of several possible trajectories leading to antisocial behavior in early childhood.
It is imperative that the developmental model of antisocial behavior be both reciprocal and transactional to handle the complexities of parent-child and environment interactions through time (Sameroff, 1990). The theory is divided into three stages of development: birth to 24 months, 24 to 42 months, and 42 months to 6 years. The division into these stages is to align the concepts with assessment periods in past follow-up studies and to identify key events that emerged in these periods. This does not imply that there are qualitatively distinct processes of transition operating at the boundaries of the periods or that the formal requirements of a stage theory will be met.
Infancy: Birth to 24 Months
It is proposed that there is one particular pathway to antisocial behavior that has its origins in the period from pregnancy through birth to age 2 years of the offspring. A plausible theoretical argument has been made by Hirschi and the attachment theorists that the most important process in this period is maternal responsiveness. The mother's aggressive, defensive, and non-conforming personality in the prenatal period sets the stage for an indifferent, neglecting response to the infant's cries, need for contact comfort, and later efforts to maintain proximity. The mother's housing and subsistence anxiety, and lack of emotional support in her social context operate as stressors to potentiate the effects of her personality.
Despite the convergence of findings demonstrating the salience of maternal responsiveness,
and given the need for a reciprocal and transactional view, it is necessary to consider how individual differences among children may exacerbate or buffer the effects of maternal non-responsiveness. In Martin's study, both the independent contributions of maternal non-responsiveness and infant demandingness, and their interaction term, assessed at 10 months of age, predicted child noncompliant behavior at 22 months. At 42 months, maternal non-responsiveness and infant demandingness were predictive of coercive child behavior. From a transactional perspective, by the end of the first year one antecedent of early coercive child behavior and inconsistent or laissez-faire parenting may be an irritable infant, or a demanding, difficult infant in the second year. It is also possible that indifferent non-responsive parenting produces the irritable, difficult, or demanding infant which then confronts the mother in later interactions. In other words, the mother may be experiencing the product of her own making.
We assume that the result of interaction between a demanding infant and a non-responsive mother is "hit or miss" dissociated interactions, as the infant makes fewer and fewer overtures, all of which are rebuffed, then avoids the mother for fear of being blocked from access (Main et al., 1985). In the latter part of this stage, 12 to 18 months, positive interactions decline in frequency, but, presumably, the mother and infant have not yet learned how to punish each other.
Equally important as the foregoing is the fact that infants in this part of the stage move from walking only while assisted to "toddling," thereby stressing the mother by moving into forbidden space and reaching for vulnerable or dangerous objects in small housing space. Then further normative increases occur in the infant's episodes of undirected anger, ability to recognize themself, and feelings of efficacy. Even the normal infant produces a negative reaction in most mothers toward the end of this period (the "terrible twos"). The negative reaction should be even stronger in indifferent, neglecting mothers, and act to produce a shift in infant attachment. Mothers of avoidant infants especially may come to view their children's behavior as demanding and difficult in this period. This perception could be at least as important as actual behavior to the prediction of later coercive cycles. Empirical support for the importance of maternal perception of infant difficulty and its relation to externalizing problem behavior at ages 4-5 has been demonstrated in two studies we have not reviewed earlier (Bates, Maslin, & Frankel, 1985; Sanson, Overklaid, Pedlow, & Prior, 1991). By 24 months the insecure-avoidant infant may have become more bold, noncompliant, and negative. It is likely that the mother not only continues to rebuff contact approaches, but is now heavily engaged in unpleasant efforts to control the infant with whom she would prefer not to interact at all.
Given the avoidant attachment, the infant fails to work with the mother cooperatively in joint tasks such as cleaning up toys or clothes on the floor. Noncompliance in such situations is only one step toward later coercive interactions, and a clear direction of movement would not occur until other normative events emerge and warp the dyadic relations further. If the mother is not successful in controlling the infant's behavior, she might become punitive and increasingly see the infant as uncontrollable by ordinary means. In a sense, her perception may be right. Presumably her task of controlling the infant's activities would be made easier if there were a secure attachment. All infants in this age range are relatively difficult to control, whether securely attached or not (Matas, Arend, & Sroufe, 1978); however, in this case, the problems of a restless, impulsive infant are superimposed on the normative invasiveness of 2-year-olds. The same intra-family processes may be seen in middle-class families, but the effects are potentially more deleterious for the male infant if his mother lacks economic and emotional resources.
Stage II: 24-42 Months
By the beginning of the second stage the basic pattern of interaction and emotional attachment between the parent and child have been formed. The child has developed an internal working model of expected responsiveness from the mother and she has developed her own expectations and standards of appropriate responsiveness concerning the child's most common reactions to her interventions. Individual differences in maternal responsiveness continue to play a salient role during the second stage. However, with the child's increased physical and cognitive maturity, the second parental factor begins to take on greater importance - parental demandingness. Some investigators with relevant data use the latter term but we will use the term "insistence" to refer to the parental effort to see that children perform developmentally appropriate tasks that require increasing social and cognitive competence. Use of the term insistence will avoid confusion with infant demandingness, and not carry the negative connotation of overly selfish and forceful parenting. Hawkins, Gottfredson, Hirschi, and Patterson, at a theoretical level, and Loeber and Dishion at an empirical level, have all alluded to a concept of insistence. Patterson, DeBarsyshe, and Ramsey (1989) note that early antisocial behavior is related directly to deficits in family management skills, as parents lacking firmness and skills in discipline are more likely to engage in coercive cycles of interaction, but also fail to tach the child skills needed in the family and school settings. In her study of early socialization practices with normal children, Baumrind (1973) found that early parental insistence (before or at age 3 1/2) was a vital component of her authoritative child-rearing topology, the parenting style most preventive of later externalizing behavior problems. It is interesting to note that the other major feature of the authoritative pattern includes parental behaviors that bear a strong resemblance to those associated with secure attachments among infants: warmth, openness, and sensitivity.
Though there are individual differences in activity level among 2-3 year-old boys, and there is some input from alternative caregivers (e.g., grandparents, older siblings) about what is deemed acceptable behavior (e.g., grandparents, older siblings), most of what children learn about regulating their own independence and aggression, and fitting into the larger family group, is determined by parents. Children do not socialize themselves; hence parents must persist in seeing that reasonable movement occurs towards this goal. Individual differences in dyads at risk for antisocial behavior outcomes take place against the background of the foregoing normative developmental changes.
The need for this new function of parental insistence is based on developmental changes that occur in all children during this period. From 24- to 42-months of age the following transformations take place that are relevant: (1) the child is able to distinguish the private self, and this increases the sense of separateness; (2) parents also come to be seen as independent and not obligated to comply with child's wishes (Ausebel's ego crisis); (3) there are decreases in undirected angry outbursts, but increases in retaliation; and (4) the child increasingly learns from parents about permitted (peers) and prohibited (parents and adults) targets of aggression. Practically speaking, the child has become physically larger and stronger, able to help with simple chores, but also capable of inflicting greater harm to siblings, peers, pets, and objects.
Whereas the secure dyad should be well on its way to establishing a "goal corrected partnership" in which it is possible to negotiate goals and make plans, the infrequent and negative physical contact between the avoidant child and mother preclude such an arrangement. Mother-infant contact occurs only when necessary (e.g., during feeding), and is characterized with a concern for reaching ends with whatever means are necessary and expedient. It could be expected that the dyad would have been transformed from one in which an avoidant attachment was dominant (attempts at access, rebuffs, avoidance to protect against rebuffs), to a dyad in which coercive cycles of interaction are emerging and bidding to become the norm. Martin's "causal" paths indicate that, from the first year forward, these cycles emerge slowly out of the cumulative effects on the infant's coerciveness of the mother's lack of involvement. There are more recent effects of the infant's noncompliance on the mother's coerciveness from the end of the second year forward. The parent-infant conflict may have become personalized and retaliatory, but there is no empirical evidence for this speculation.
By the end of this stage, mothers involved in dyads who follow this hypothesized sequence would be inconsistent and/or inflexible administrators of discipline. Those that initially persisted in their requests for compliance may not be consistent in enforcing these demands as their children gradually show that such efforts have little effect. Alternatively, defiant, oppositional children may only receive demands for compliance from an authoritarian parent, and the interactions might be devoid of any affective bond. In both cases, it is reasoned that the children would be noncompliant, negative, and unenthusiastic towards their parents. The children might not trust the parents to adequately care for them or their needs. Thus, even if such mothers are able to maintain a consistent firmness in dealing with their children's behavior, they are more likely to meet greater resistance than with the average child, who would have more to lose by maintaining noncompliant behavior (i.e., loss of love). Furthermore, it is more than likely that these parents perceive their children as hostile, based on the original perception of the child and the child's increasing noncompliant behavior. The children, in turn, come to interpret neutral or even friendly behavior as hostile. The mothers' use of ineffectual disciplinary techniques fail to produce the desired outcome, and thus reinforces a fatalistic attitude, thus setting up a self-fulfilling prophecy: he always was, and always will be, a bad child.
Stage III: 42-60 Months
During the third stage of the model, the dyadic processes between parent and child become consolidated and spread to other family members. Responsivity and insistence have become well established. Other relevant normative changes involve the child's increasing ability to function both outside and inside the home, and to handle the interpersonal and cognitive problems encountered. It is during the early part of this stage that children begin taking what they have learned in the home about aggressive behavior and trying it out in new settings such as preschool. The extrinsic-intrinsic motivational system emerges: children who have experienced parental rewards that enhance feelings of self control emerge with an intrinsic motivational orientation. On the other hand, children who have been rewarded as a means to control their behavior tend to only pursue a rewarded activity in the presence of the parents, and do not independently pursue activities that are complex and challenging.
It is during Stage III that responsivity and insistence must be well established by the parent to prevent an increase in problem behavior. As mentioned previously, with the foundation of a secure attachment, parental requests for appropriate behavior are responded to increasingly with compliance and internalization. On the other hand, simply having a secure attachment with the caregiver does not insure that a child will not show aggressive behavior. The oppositional behavior may originate, and be reinforced initially among siblings or peers. If the parent is not alert in prohibiting aggression in its early manifestations, or does so in an inconsistent manner, it may evolve into a power struggle, and the aggression would be rewarded inadvertently. Both Baumrind (1973) and Block (1971) found that the effects of a permissive child-rearing pattern, marked by warmth without control, predicted later externalizing behavior outcomes that were more deleterious than outcomes from the authoritarian child-rearing pattern, characterized by demandingness without responsiveness.
Against the background of the foregoing interaction processes and normative changes in families and children, the individual differences associated with later antisocial behavior are hypothesized as follows. In general, continuity of conflict is the rule between Stages II and III: Parents maintain the coercive intervention strategies used in Stage II, but the conflict is more intense, and now there is vacillation between ignoring rule violations and employing harsh or threatening punishments. There is also continuity in the mother's image of her child as hostile, noncompliant, and aggressive, and much more difficult to handle at this stage. At the same time, the child is applying to the preschool setting the pattern of permitted aggression at home, and the result in this case is not complaints from teachers, but threats by school officials to remove the child from the school unless behavior changes. The parents, accustomed to tolerating hostile oppositional behavior, are surprised at the threats by school officials. Parents are often surprised and angered by the complaints of teachers. Even though they see some of the same behaviors at home, they do not take action since the parent-child equilibrium can be described as homeostatic from a family systems perspective (Minuchin, 1974). Although contact between the child and parent continues to be infrequent, distant, and aversive, the dyad has settled into its patterns with little dissonance concerning the need for change.
More important for the development of later antisocial behavior, there is little internalization of parental and societal standards even when compliance is secured. Rewards are infrequent and used only as a means of controlling behavior. Children who have an avoidant relationship with their caregivers are likely to comply only because of perceived threats to their freedom or physical safety. All they assimilate from their infrequent conformity to authority is an extrinsic motivational system. They do things because they have to and the parent is standing over them. They do not independently pursue activities that are complex and challenging.
Summary
In view of the increased interest in a developmental approach to psychopathology, and mounting evidence of the importance of parent-child interaction in the etiology of antisocial behavior, the following questions were posed for this review. What theories of parent-child relationships and family management techniques are available, how developmental are they, how specific are they relative to parent and child behaviors involved, and how well do they cover the period in which antisocial behavior develops? Six theories have some developmental features but the attachment theories (Sroufe & Egeland; Greenberg) and two social learning theories (Patterson and Martin) are most clearly developmental in that they postulate reciprocal interactions of parent and child, transactions, and transformations in the form of normative changes in the child or changes in family processes. The social learning theories of Patterson and Martin are most specific, microanalytic in fact, as to the interaction processes involved, but the attachment theories at least specify kinds of behavior involved and also do not rely on traits or types of influence as their units of analysis.
It is worth noting that the social learning theories have moved beyond the "laundry list" stage of conceptualization in which long lists of relevant variables are offered but with no specification of mechanisms or processes that align these variables developmentally (see Jessor, 1984). Patterson contends that many demographic variables only affect development of antisocial behavior to the extent that they produce perturbations in family management. Now it appears that the infant can produce these perturbations. Martin found that the interaction of infant demandingness and maternal responsiveness was related to the development of later coercive cycles, above and beyond the effects of demandingness and responsiveness alone.
Conceptualization is most weak and overly general between late infancy and the preschool years. This gap makes it difficult to link attachment and social learning theories, both of which have driven a large number of studies. A bridging theory is offered to link the two sets of theories in the critical period involved.
References
Ackerman, N. (1954). Interpersonal disturbances in the family: Some unsolved problems in psychotherapy. Psychiatry, 17, 359-368.
Bates, J.E., Maslin, C.A., & Frankel, K.A. (1985). Attachment security, mother-child interaction, and temperament as predictors of behavior-problem ratings at age three years. In I. Bretherton & E. Waters (Eds.), Monographs of the Society for Research in Child Development, 50, Nos. 1-2, 167-193.
Baumrind, D. (1973). The development of instrumental competence through socialization. In A. Pick (Ed.), Minnesota Symposia on Child Psychology (Vol. 7, pp. 3-46). Minneapolis: University of Minnesota Press.
Bell, R.Q. & Harper, L.V. (1977) Child effects on adults. Hillsdale, New Jersey: Lawrence Earlbaum Assoc.
Bell, R.Q., & Chapman, M. (1986). Child effects in studies using experimental or brief longitudinal approaches to socialization. Developmental Psychology, 22, 595-603.
Bell, R.Q. (1986). Age specific manifestations in changing psychosocial risk. In D.C. Farren & J.D. McKinney (Eds.), The concept of risk in intellectual and psychosocial development (pp. 169-185). New York: Academic Press.
Block, J. (1971). Lives through time. Berkeley, CA: Brancroft Books, 40.
Bowlby, J. (1980). Attachment and loss: Vol. 3, Loss, sadness and depression. New York: Basic Books.
Bowlby, J. (1973). Attachment and loss: Vol. II Separation: Anxiety and anger. London: Hogarth Press.
Bowlby, J. (1969). Attachment. New York: Basic Books.
Egeland, B., & Farber, E.A. (1984). Infant-mother attachment: Factors related to its development and changes over time. Child Development, 55, 753-771.
Erickson, M.F., Sroufe, L.A., & Egeland B. (1985). The relationship between quality of attachment and behavior problems in preschool in a high-risk sample. In I. Bretherton & E. Waters (Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50, Nos. 1-2, 147-167.
Fagot, B.I,. & Kavanagh, K. (1990). The prediction of antisocial behavior from avoidant attachment classifications. Child Development, 62, 864-873.
Greenberg, M.T., & Speltz, M.L. (1988). Contributions of attachment theory to the understanding of conduct problems during the preschool years. J. Belsky & T. Negworski (Eds.), Clinical implications of attachment (pp. 177-218). Hillsdale, New Jersey: Earlbaum.
Glueck, S. (1950). Unraveling juvenile delinquency. S. Glueck & E. Glueck (Eds.), New York: Commonwealth Fund.
Gottfredson, M.R., & Hirschi, T. (1990). A general theory of crime. Stanford, California: Stanford M. Green (Ed.).
Hawkins, J.D., Lishner, J.M., Catalano, G.E., Howard, M.A. (1986). Childhood predictors of adolescent substance abuse: Toward an empirically grounded theory. Journal of Children in Contemporary Society, 8, 11-40.
Hirschi, T. (1969). Causes of Delinquency. Berkeley, CA: University of California.
Jackson, D.D. (1957). The question of family homeostasis. The Psychiatric Quarterly Supplement, 31, 79-90.
Jessor, R. (1984). Adolescent development and behavioral health. In J.D. Matarazzo, S.M. Weiss, J.A. Herd, N.E. Miller, & S.M. Weiss (Eds.), Behavioral health: A handbook of health enhancement and disease prevention (pp. 69-90). New York: John Wiley & Sons.
Lamb, M.E., Thompson, R.A., Gardner, W.P., Charnov, E. L. & Esler, D. (1984). Security of infantile attachment as assessed in the "strange situation": Its study and biological interpretation. The Behavioral and Brain Sciences, 7, 127-171.
Lepper, M.R. (1982). Social control processes, attributions of motivation, and the internalization of social values. In E.T. Higgins, D.N. Ruble, & W.W. Hartup (Eds.), Social cognition and social behavior: Developmental perspectives (pp. 294-332). Cambridge, England: Cambridge University Press.
Lewis, M., Feiring, C., McGuffog, C. & Jaskir, J. (1984). Predicting psychopathology in six-year-olds from early social relations. Child Development, 55, 123-136.
Loeber, R. & Dishion, T.J. (1983). Early predictors of male delinquency: A review. Psychological Bulletin, 94, 68-99.
Loeber, R., & LeBlanc, M. (1990). Towards a developmental criminology. In M. Tonry & N. Morris (Eds.), Crime and justice: A review of research (Vol. 12, pp. 375-473). Chicago: University of Chicago Press.
Loeber, R. & Stouthamer-Loeber, M. (1986). Family factors as correlates and predictors of juvenile conduct problems and delinquency. In M. Tonry and N. Morris (Eds.) Crime and Justice: A review of research (Vol. 7, pp. 29-149). Chicago: University of Chicago Press.
Lytton, H. (1990). Child and parent effects in boys' conduct disorder: A reinterpretation. Developmental Psychology, 26, 683-697.
Maccoby, E.E., & Martin, J.M. (1983). Socialization in the context of the family: Parent-child interaction. In E.M. Hetherington (Ed.), Socialization, personality and social development (Vol IV, Chap. 1, pp. 1-102). New York: Wiley.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adulthood: A move to the level of representation. In I. Bretherton & E. Waters (Eds.), Growing points of attachment theory and research. Monographs of the Society for Research in Child Development, 50, (1-2, Serial No. 209, pp. 66-104).
Martin, J. (1981). A longitudinal study of the consequences of early mother-infant interaction: A microanalytic approach. Monographs of the Society for Research in Child Development, 46, (3, Serial No. 190).
Marvin, R.S. (1977). An ethological-cognitive model for the attenuation of mother-child attachment behavior. In T.M. Alloway, L. Crames, & P. Pliner (Eds.), Advances in the study of communication and affect: Vol. 3. The development of social attachments (pp. 25-60). New York: Plenum.
Marvin, R.S., & Greenberg, M.T. (1982). Reactions of preschool children to adult strangers: A behavioral systems approach. Child Development, 53, 481-490.
Matas, L., Arend, R.A., & Sroufe, L.A. (1978). Continuity of adaptation in the second year: The relationship between quality of attachment and later competence. Child Development, 49, 547-556.
McCord, W., McCord, J., & Zola, I.K. (1959). Origins of crime. New York: Columbia University Press.
McCord, J. (1978). A thirty-year follow-up of treatment effects. American Psychologist, 33,
284-289.
Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press.
Patterson, G.R., Capaldi, D. & Bank, L. (1991). The development and treatment of childhood aggression. In D. Pepler, & R.K. Rubin (Eds.), The development and treatment of childhood aggression (pp. 139-168). Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Patterson, G.R., DeBarsyshe, B.D., & Ramsey, E. (1989). A developmental perspective on antisocial behavior. American Psychologist, 44, 329-335.
Reid, J.B., & Patterson, G.R. (1991). The development of antisocial behavior patterns in childhood and adolescence. European Journal of Personality. NEED PP, VOL.
Renken, B. Egeland, B., Marvinney, D., Mangelsdorf, S., & Sroufe, A. (1989). Early childhood antecedents of aggression and passive-withdrawal in early elementary school. Journal of Personality, 57, 257-281.
Robins, L.N. (1966). Deviant children grow up. Baltimore, MD: Williams & Wilkins.
Sameroff, A.J. (1990). Prevention of developmental psychopathology using the transactional model: Perspectives on host, risk agent, and environment interactions. Paper presented at the Conference on The Present Status and Future Needs of Research on Prevention of mental Disorders, Washington D.C.
Sameroff, A.J., & Chandler, M.J. (1975). Reproductive risk and the continuum of caretaking casualty. In F.D. Horowitz, M. Hetherington, S. Scarr-Salapatek, & G. Siegel (Eds.), Review of child development research (Vol. 4, pp. 187-204). Chicago: University of Chicago Press.
Sameroff, A.J., Seifer, R., & Zax, M. (1982). Early development and children at risk for emotional disorder. Monographs of the Society for Research in Child Development, 47, (7, Serial No. 199).
Sanson, A., Oberklaid, F., Pedlow, R., & Prior, M. (1991). Risk indicators: Assessment of infancy predictors of pre-school behavioural maladjustment. Journal of Child Psychology and Psychiatry, 32, 609-626.
Sroufe, L.A. (1983). Infant-caregiver attachment and patterns of adaptation in pre-school: The roots of maladaptation and competence. In M. Perlmutter (Ed.), Minnesota symposium in child psychology (Vol. 16, pp., 41-81). Hillsdale, New Jersey: Erlbaum.
Stayton D.J., Hogan, R., & Ainsworth, M.D.S. (1971). Infant obedience and maternal behavior: the origins of socialization reconsidered. Child Development, 42, 1057-1069.
Thelen, E., Kelso, J.A.S., & Fogel, A. (1987). Self-organizing systems and infant motor development, Developmental Review, 7, 39-65.
West, D.J. (1969). Present conduct and future delinquency. London: Heinemann.
Footnotes
Requests for reprints should be sent to the first author at the Department of Psychology,
Clinical Psychology Center, 604 OEH, 4015 O'Hara Street, University of Pittsburgh,
Pittsburgh, PA, 15260, (412) 624-8790. Portions of this paper were presented in Shaw, D.S., &
Bell, R.Q. (1992, May), Developmental precursors of antisocial behavior, at the Conference for
Life History Research, Coercion and punishment in long-term perspectives, Philadelphia, PA.