Implications of Research on Infant Development for Psychodynamic Theory and Practice

CHARLES H. ZEANAH, M.D., THOMAS F. ANDERS, M.D., RONALD SEIFER, Ph.D., AND DANIEL N. STERN, M.D.

Abstract. Recent research on infant development is reviewed to consider its implications for psychodynamic theory and practice. To address the question of the importance of early experiences for development, research on continuities and discontinuities in development, temperament, motivational systems in infancy, affect develop­ ment and regulation, development of the sense of self, and infant-caregiver attachment are reviewed. Two major implications emerge, both emphasizing the need for more complexities in our conceptualizations. First, research on infant development underscores the importance of context in development and cautions about the limits of reductionistic thinking and theories. Second a major paradigmatic shift away from the fixation-regression model of psychopathology and development is indicated. A new model that better fits available data is proposed instead. In this continuous construction model, there is no need for regression, and ontogenetic origins ofpsychopathology are no longer necessarily tied to specific critical or sensitive periods in development. Implications for psychody­ namic treatment are also described. J. Am. Acad. Child Adolesc. Psychiatry, 1989, 28, 5:657-668. Key Words: psychodynamic theory, infant development, sense of self, fixation-regression.

Freud (1940) described the infant's relationship to his or ded. From the standpoint of infant development, this theme her mother as "unique, without parallel, established unalter­ of contexts highlights the importance of early infant-caregiver ably for a whole lifetime as the first and strongest love object relationship experiences. and as the prototype for all later love relations" (p. 188). Research on infant development has led to the second Following in this tradition, most other psychodynamic theo­ theme, that is, supplanting drive theory's fixation-regression rists have accorded a vitally important role to infancy and model with a continuous construction model of development early relationship experiences. In the last 25 years, an increas­ and psychopathology. Traditionally, psychodynamic theory ing number of methodologically rigorous studies on infant has considered relationship experiences to be organized by development have yielded results that contribute significantly the oral, anal, and phallic libidinal stages or modes of relating. to psychodynamic assertions about the nature and conse­ Psychopathology is understood to derive from regression to quences of early experiences. Although many important ques­ fixation points, the latter resulting from constitutional vul­ tions remain, and controversies abound, a coherent picture is nerabilities and/or from infantile traumas within certain crit­ emerging. In this selective review of research on infant devel­ ical or sensitive periods. This fixation-regression model of opment, we discuss two major themes germane to psychody­ psychopathology also guides treatment. Psychoanalysis or namic theory and practice. psychoanalytic psychotherapy proscribes gradually uncover­ The first theme concerns the importance of context in ing the original trauma or fixation point so that it may be development and emphasizes the continuing and mutually worked through. In this conceptualization, the ontogenetic reciprocal transactions between individual and environment. origin of psychopathology arises from the particular libidinal Recognizing the multi-determined interactive complexities of stage of development in which there is a trauma or a consti­ development and psychopathology requires moving beyond tutional weakness. reductionistic thinking and theory building. This perspective In the continuous construction model, patterns of internal is particularly important to emphasize during the current subjective experience and patterns of relating to others are period of dramatic advances by neuroscientists and molecular derived from past relationship experiences but are continu­ biologists. What has been learned about development from ously operating in the present. The model is "continuous" in infancy research strongly suggeststhat searching for the causes two senses. First, it alludes to dynamic, ongoing transactions of disordered behavior and personality organization in genetic between individual and environment throughout develop­ markers and biochemical defects will never provide the com­ ment. Second, continuous refers to the overall coherence of plete picture. Whatever contributions the synaptic cleft and an individual's sense of self and personal history, as well as the receptor make to psychopathology are more meaningfully the coherence of his or her relationship patterns. This coher­ considered within the larger contexts in which they are imbed- ence implies both repetition of previous patterns and a tend­ ency to resist change. The shift from a fixation-regression model to a continuous Accepted April 17. 1989. From the Division 0/ Child and Adolescent Psychiatry. Brown construction model has important implications for the onto­ University Program in Medicine. genetic origins of psychopathology and for the conduct of Dr. Zeanah is supported in part by a Research Scientist Develop­ psychotherapy. The most important difference in the contin­ ment Award/rom NIMH (MH00691). uous construction model and the fixation-regression model is Reprint requests to Dr. Zeanah, Women and Infants' Hospital. 101 that the former does not specifythe point of origin of various Dudley Street. Providence. RI02860. 0890-8567/89/2805-Q657$02.00/0© 1989 by the American Acad­ forms of psychopathology or link them to particular devel­ emy of Child and Adolescent Psychiatry. opmental phases. Instead, it leaves the question of putative 657 658 ZEANAH ET AL. origin open, allowing each issue's origin to emerge from each becoming more prolonged and shifted to the night. Many patient's own particular lifestory. Since the pattern is ongoing, infants begin for the first time to sleep through the night. "ingression" (rather than regression) into salient clinical issues Enhanced cognitive abilities are evident by more rapid learn­ in the here and now occurs through the transference relation­ ing in habituation, classical conditioning, and operant con­ ship. ditioning. Even more striking for parents are the social and Since both the fixation-regression model and the continu­ affective changes. Soon after crying or "colicky" behavior ous construction model imply some degree of continuity begins to diminish, infants begin to make prolonged eye-to­ between early and later experiences, the concept of develop­ eye contact and smile and coo responsively in the presence of mental continuities and discontinuities is reviewed. Then how a human face (Emde and Harmon, 1972; Emde et al., 1976). continuity relates to empirical findings on infant tempera­ The next period of reorganization occurs at around 7 to 9 ment, motivational systems, and affect development is dis­ months of age. Biological changes permit postural and loco­ cussed. Next, because the sense of self integrates these do­ motor advances. In the cognitive domain, infants develop a mains of development and is closely linked to subjective rudimentary understanding of means-ends differentiaton, experience, continuities and discontinuities in the early de­ along with intentionality, object permanence, and a sense of velopment of self are reviewed. Specifically, two major theo­ anticipation. Affectively, the consolidation of these advances retical models of self development that have been drawn from is accompanied by the onset of a specific attachment to a infancy research are reviewed and their differing psychody­ primary caregiver, manifest in part by separation reactions namic implications are considered. In keeping with the rec­ and stranger wariness. ognition that continuity in infant development is even clearer At 12 to 13 months, more changes appear. Cognitively, at the level of relationship patterns, theoretical and empirical infants have a new appreciation of the independence of enti­ work on patterns of infant-caregiver attachment relationships ties in the world. Mobility is greatly enhanced by the infant's and their implications for the importance of early experience beginning to walk, further contributing to a new sense of the are described. In the concluding section, the implications of infant as a more autonomous individual. Affective advances these research results for conceptualizing psychopathology are evident by the appearance of social referencing to resolve and conducting psychotherapy are highlighted. uncertainty (Campos and Stenburg, 1981). Affects begin to be used instrumentally by the infant (K1innert et al., 1984), Continuities and Discontinuities in Development and affect attunement by caregivers provides the infant with Belief in the importance of early "traumatic" experiences a richer experience of empathy (Stern, 1985b). in the formative years of infancy and an interest in identifying The final transformation in the first 2 years occurs at 18 to individuals at risk for later psychiatric disorder has led inves­ 20 months. Symbolic representation dramatically advances, tigators to search intensively for significant continuities in and language becomes increasingly preferred as a means of development. The medical model of psychopathology led communication. For the first time, infants understand con­ early psychoanalytic theorists to predict that psychological sistently that words are an agreed upon means of designating traumas and biological propensities lead to predictable se­ objects and later events. Self-awareness, gender recognition, quellae and consequences. Contrary to expectations, one of and the capacity for self-reflection first become evident. Ad­ the major results of the search for continuities in behavior vances in peer relatedness soon follow these other develop­ has been the recognition that discontinuities in early devel­ ments. opment are far more readily apparent than continuities (Emde Each of these developmental transformations leads to qual­ and Harmon, 1984; Rutter, 1987). This, coupled with evi­ itative shifts in biological, cognitive, affective, and social dence of adequate coping in some resilient children and adults organization, changing both how the infant experiences the despite adverse early experiences, has led some to ascribe world and how others experience the infant (Stern, 1985a). little, if any, significance to experience in the early years Recognition of these developmental discontinuities suggests (Kagan, 1984a; Clarke and Clarke, 1976). that simple links between early and later behaviors are un­ An initial conundrum for developmentalists interested in likely. Development means inevitable changes in both the continuities has been the widespread recognition that major individual and the environment. The widely accepted trans­ periods of developmental reorganization occur during the first actional model of development (Sameroff and Chandler 1975) two years of life. These biobehaviora/ shifts are rather discrete describes the process by which the infant is changed by epochs during which biological, cognitive, affective, and social environmental influences even as he or she influences and characteristics of the infant reorganize and subsequently changes that environment. The continuing, mutually recip­ emerge as qualitatively new capacities of greater complexity rocal relationship between infant and environment requires (Emde et aI., 1976; Emde, 1985). These changes have been an understanding of continuity as a product of this increas­ widely recognized and extensively reviewedelsewhere (McCall ingly complex process and underscores the importance of et al., 1977; Kagan, 1984b; Emde, 1985; Stern, 1985a). The considering infant development within its environmental con­ major features of the biobehavioral shifts are summarized text. Research reviewed below on temperament, motivational briefly below. systems, and affect development addresses some of the com­ Around 2 to 3 months after birth, a variety of behavioral plexities of early development as they effect continuity as a changes may be observed. Although the total amount of sleep basic psychodynamic principle. In addition, the review draws does not change dramatically, a diurnal cycle of sleep/wake attention to important components of subjective experience appears, with brief waking periods becoming consolidated that provide the foundation for an individual's sense of per­ into longer daytime bouts of wakefulness and sleep periods sonal, historical continuity. PSYCHODYNAMICS: INFANT DEVELOPMENT RESEARCH 659

Temperament though there are scattered findings relating both parental reports and laboratory assessments of temperament to some Investigators of infant temperament initially hoped to de­ clinical conditions (Rothbart and Goldsmith, 1985; Carey, scribe early, stable manifestations of personality. Following 1986; Rosenbaum et aI., 1988). These links are greatly the approach of Thomas, Chess, and colleagues (Thomas et strengthened when the infant's temperament is considered in al., 1963), most investigators and theorists have considered context (Graham et aI., 1973; Cameron, 1978). Wolkind and temperament to be the style by which individuals behave in DeSalis (1982), for instance, found that 4-month-old infants different situations over time, attempting to capture the es­ with difficult temperaments were significantly more likely to sence of those behaviors that comprise personality. Temper­ be behavior disordered at age 3 years if their mothers were amental dimensions refer to behavioral tendencies rather than clinically depressed. discrete behavioral acts (Goldsmith et aI., 1987). Typically, All of these results have led to interest in considering temperament has been aligned with the notion of personality temperament in the context of the parent-child relationship. traits rooted in an individual 's biology (Goldsmith et aI., Thomas, Chess, and colleagues (Thomas et aI., 1963, 1968) 1987). originally formulated the "goodness-of-fit" construct, empha­ Research in behavioral genetics have examined parent­ sizingthe interaction between parental responses and differing child, sibling, and twin concordances, measured by parental infant temperamental types in the production of disordered reports. Overall, results indicate a significant contribution of behavior. Seifer and Sameroff(I986) expanded this model to genetic factors to temperament (Plomin, 1987). Several di­ integrate more systematically the contextual factors associated mensions of temperament appear to have genetic contribu­ with parental responses to and interpretation of differing tions, including emotionality, activity, and sociability (Buss behavioral styles in young children. As a result, their model and Plomin , 1984). Interestingly, stability of inherited traits highlights the importance of the subjective experiences of may actually increase with increasing age, pointing to the parent and infant and emphasizes the importance of context complexity of gene-environment interaction (Scarr and Kidd, in understanding early development and its disorders. Re­ 1983; Plomin, 1987). In a similar manner, studies of behav­ search on infant temperament underscores the fallacy of ioral inhibition (shyness) have demonstrated continuity in theoretical models that attempt to explain development and selected extreme samples from 2 to 6 years of age. A number psychopathology without sufficient attention to context. of biological correlates of this characteristic also have been demonstrated (Kagan et aI., 1988). Motivational Systems The major methodologic problem for temperament re­ search has been concern about the adequacy of parental As a result of infancy research, motivational systems in reports for assessing childrens' temperaments. Parents have infancy have been expanded beyond the libidinal and aggres­ been shown to have systematic biases in their ratings based sive drives proposed by Freud. One of the domains of moti­ on their social class, race, anxiety level, and mental health vation that recent infancy research has described compellingly status (Sameroff et al., 1982; Vaughn et aI., 1987). To a is curiosity. From the moment of birth, infants actively and degree, parents also develop preconceived notions even before preemptorily seek stimulation, especially social stimulation their children are born as to what their temperaments will be (Stem, 1977; Anders and Zeanah, 1984). Since Wolff's (1966) (Zeanah et aI., 1985, 1986a; Mebert and Kalinowsky, 1986). seminal work on the importance of the state of alert inactivity Mother-father-observer agreements about temperament are for learning, a host of investigations using paired comparison only moderate or less (Hubert et aI., 1982; Field and Green­ and habituation paradigms have capitalized on infant prefer­ berg, 1982; Zeanah et aI., 1986b), perhaps reflecting the ences for novelty to discover other infant capacities. Stem generally low (about 0.25) agreement among different inform­ (1985a) cites evidence from such paired comparison para­ ants who have different roles with a particular child (Achen­ digms indicating that under conditions of novelty, even young bach et aI., 1987). The basic question from these studies infants would rather look than eat, underscoring the impor­ remains, is continuity in the infant's behavior or in the tance of curiosity as a basic motivational system. Infants perception of the informant? possess inborn biases and preferences for certain types of Clinically, the most important construct to emerge from stimulation, along with a tendency to explore, to form and temperament research is that of infant difficulty (Bates, 1980), test hypotheses about what is occurring in the world (Brunner, although there has been considerable debate about its meaning 1977; Kagan et aI., 1978; Lamb and Sherrod, 1981; Lipsitt, and measurement (Bates, 1980; Thomas et aI., 1982; Bates, 1983). 1983; Carey, 1986; Zeanah et al., 1986c). Adverse tempera­ Another closely related area of research on basic motiva­ ment characteristics in infancy, if stable, provide an oppor­ tional systems is mastery motivation. Research on mastery tunity for early identification of individuals at risk for subse­ motivation was inspired by White's (1959) theory of effect­ quent psychopathology. The problem has been that evidence ance motivation, a relatively early divergence from accepted for continuity of temperamental traits is equivocal at best, psychoanalytic theory on the nature of drives. White believed especially when measured in infancy (Hubert et aI., 1982). that individuals are intrinsically motivated to explore and to Despite these conceptual differences and shortcomings in master their environmental contexts. His position was con­ assessment, temperament remains a compelling if frustrat­ gruent with Hunt's (1965) view of infants as motivated, active ingly elusive construct. From the clinical standpoint, the constructors of their environments. Following in this tradi­ major point about temperament is that direct links between tion, Harter (1978) developed a theoretical model that links temperament types and psychiatric disorders are limited, al- environmental consequences of mastery behavior with indi- 660 ZEANAH ET AL. vidual differences in self-esteem. Specifically, contexts that predicts mastery behavior 8 months later. Maternal control encourage and reinforce attempts at mastery are believed to stylesand sensitivity also affected mastery behavior of l-year­ promote positive feelings of self-efficacy and internal percep­ old infants (Grolnick et al., 1984). Following these leads, tions of control. Seifer et al. (in press) have developed a model of social White, Hunt, and Harter emphasized the universal aspects competence in toddlers that emphasizes the integration of of mastery motivation as a behavioral characteristic of the security of attachment, mastery motivation, affect expression, species. Yarrow and colleagues (Yarrow et aI., 1983) began a modulation and self-control, and task success. series of studies of infants in the first year of life to determine The most important psychodynamic implication of these individual differences in the degree and manner in which research findings concerns the maintenance of self-esteem. mastery motivation was expressed. Their strategy for meas­ Research in this area points to infants' motivation by curiosity uring mastery motivation was to present a series of toys that and the desire to explore, by the urge to master, and by the emphasize production of effects, sensorimotor practice, or desire for competence to behave in ways that maintain self­ problem solving to infants, and to assess the degree to which esteem. In other words, evidence suggestsa variety of behav­ the infants explored and mastered the situation with a mini­ iors that are more usefully explained by other motivational mum of adult guidance. Clear individual differences in atten­ systems rather than as discharges of excess psychic energy. tion, persistence, exploration, and task success were demon­ And maladaptive behaviors can be explained in terms other strated in these studies (Yarrow et al., 1983; Morgan and than compromise solutions to conflict situations. Harmon, 1984). Further, these differences have been related to developmental status, handicapping conditions, and inter­ Affect Development and Regulation actions with parents (Jennings et al., 1988; Macturk et aI., 1985; Yarrow et aI., 1982, 1984). In addition, it appears that Research on infants and adult-infant relationships indicates goal-directed behavior, rather than task success, is related to that throughout development, emotions regulate behavior, self-efficacyas manifest in contingent positive affect displays internal processes, and social interactions (Barrett and Cam­ (MacTurk et al., 1987). pos, 1987).For example, much as Freud speculated, fear helps Some theoreticians, particularly in the field of self psychol­ the individual avoid dangerous situations, activate the fight­ ogy, have asserted that the overarching motivational system flight response, and alert others to the presence of danger. driving behavior is a feeling of competence (Basch, 1988). In This research also supports traditional psychodynamic theory this line of reasoning, competence subsumes curiosity, mas­ about the role of emotions in personality formation and about tery, effectance, and even sex and aggression as a superordi­ their importance in clinical treatment (Emde, 1988a, b). nate construct. Competent behavior is the external goal and The impetus for research on affect development comes self-esteem is the internal goal in this important motivational from demonstrations of ontogenetic changes in facial expres­ system. sions and the recognition of seven discrete emotions-joy, Sroufe and colleagues have inspired a growing body of sadness, anger, fear, surprise, disgust, and interest-in exten­ research on the development of social competence in infants sive cross-cultural studies (Izard, 1971; Ekman et al., 1972). and toddlers (Sroufe and Waters, 1977; Matas et al., 1978; The implication of this work is that parents and infants are Waters and Sroufe, 1983). Two major points are emphasized biologically predisposed to display and to interpret certain in their work. First, social competence is defined as the flexible emotional expressions in one another and to interact on the and effective use of available personal, social, and physical basis of these interpretations. resources to achieve goals (Waters and Sroufe, 1983).Second, Further work demonstrates that facial expressions of emo­ infant behaviors are examined in terms of their organization tions show regularity in their form and time of emergence. (Sroufe and Waters, 1977);that is, patterns of behavior instead Both parents and objective observers reliably recognize facial of isolated behaviors are assessed. Crucial is the appreciation, expressions in infants in the first year of life (Emde, 1985, familiar to clinicians, that individual infant behaviors have 1988a). In fact, interest, disgust, and physical distress are different meanings in different contexts and at different de­ readily demonstrable at birth and have specifiable elicitors velopmental periods. Behaviors are considered intimately tied and predictable interactional consequences (Izard and Mala­ to the important relationships ofthe child in various contexts testa, 1987). In the first 7 months of life, the maturation of and in different developmental periods. Meaningfully orga­ surprise, sadness, anger, and fear appear to be primarily nized patterns ofbehavior are more stable and predictive than biologically programmed, although social experiences may individual behaviors alone (Sroufe and Waters, 1977;Waters, have great influence as early elicitors and contingent reinfor­ 1978; Sroufe, 1983). cers of these actions. Two year olds' effectiveuse of socialsupport is an important Emotional expressions are especially important motivators ingredient in their ability to function competently in problem and regulators of social behavior. In early infancy, caregiver solving tasks. This includes seeking help when needed and at responses are differentially influenced by infant expressions times rebuking unnecessary interference by adults. Individual of emotions (Malatesta, 1981), and infants are in turn influ­ differences in dyadic interaction in this regard are readily enced by the caregiver's affective expressions (Tronick et al., demonstrable (Matas et al., 1978) and are predictive of later 1986). With the maturation of higher order central nervous peer relationships (Sroufe, 1983). system inhibitory mechanisms and the development of more Recent investigations have taken a more integrative ap­ complex cognitive abilities, emotional expressions become proach to studying mastery and competence in infancy. Frodi increasingly subject to control. They also become increasingly et al. (1985) have found that security of attachment at I year useful as a means of regulating the behavior of others. Al- PSYCHODYNAMICS: INFANT DEVELOPMENT RESEARCH 661 though many caregivers already may interpret infant affect theories about the early development of the sense of self. expressions as intentional in neonates (Kaye, 1982), they still Rather than reviewing in detail the voluminous research with recognize a qualitative change towards the latter part of the infants bearing on self development, we present a review first year as they begin to perceive infants using their emo­ instead of the two major conflicting theoretical perspectives, tional expressions instrumentally to obtain desired goals Mahler's (Mahler et al., 1975) theory of separation-individua­ (Klinnert et aI., 1984). tion and Stern's (l985a) theory of self development. These The importance of the emotional availability of caregivers theories share a central focus on the infant's sense of self and to their infants has been stressed by many psychodynamic on the overarching clinical issues of being alone and being theorists and investigators (Bowlby, 1969/1982; Mahler et al. with others as fundamental human concerns. Nevertheless, 1975; Ainsworth et al., 1978; Emde and Easterbrooks, 1985; their implications for psychodynamic theory and practice lead Sroufe, 1988). Recent research with l-year-olds has demon­ us in different directions. The theory of separation-individua­ strated that in situations of uncertainty, infants look to their tion is fully compatible with drive theory's fixation-regression caregivers for help in evaluating their emotional reaction to model, but the self development theory of Stern illustrates the the uncertainty. This form of emotional signaling has been continuous construction model. As a result, they necessarily termed social referencing (Campos and Stenberg, 1981; Klin­ provide different and incompatible views of the course of self nert et aI., 1982). Social referencing, which occurs at a time development, the origins of psychopathology, and the strate­ when the infant might have more difficulty with evaluating gies of psychotherapy. the safety and consequences of his or her own actions and the Compatible with empirical findings that self-awareness de­ new meanings of a variety of environmental events, appears velops at about 18 months (Lewis and Brooks-Gunn , 1979) to be an important precursor of real self-awareness (Ernde, is the influential theoretical work on self development that 1985). began with Mahler and her colleagues (Mahler et aI., 1975). Infants begin to label their own feeling states towards the Based on extensive, longitudinal observations of a small sam­ end of the second year oflife (Bretherton and Beeghly-Smith, ple of middle-class mothers and children . together with clini­ 1982). Together with capacities for self-awareness, this period cal experiences with older psychotic children, Mahler and her marks the initial appearance of an observing ego. The child colleagues elaborated a widely known theory of self develop­ can have one feeling and simultaneously think about another ment. They proposed that an infant's psychological birth is a feeling and deal with it symbolically, opening up new possi­ gradual process that extends throughout the first 2 years after bilities for complexity and for conflict between internal ex­ physical birth. The gradual discovery of self-awareness and perience and external expression. There is substantial evi­ independence from the mother was described as the process dence that infants can learn to deny emotional expression of of separation-individuation. For the first 2 months, infants ongoing feelings, thus dissociating expression and feelings are in a normal autistic phase. in which physiological proc­ (Izard Barrett and Campos, 1987; Izard and Maletesta, 1987). esses are dominant over psychological processes. From 2 to 6 The significance of continuity of emotional development months, in the normal symbiotic phase, the infant "behaves has been stressed most explicitly by Emde (1985). He suggests and functions as though he and his mother were an omnipo­ that it is an individual's emotional experiences that give a tent system-adual unity within a common boundary" (Mah­ sense of identity throughout development. Beingable to access ler et al., 1975, p. 44). our own feelings consistently gives us a sense of familiarity Following this phase, the process of separation-individua­ about who we are despite many changes over time. Further, tion begins. During the initial phase of differentiation, the because this "affective core" is rooted in biology, it provides infant "hatches" out of the "symbiotic orbit" and begins to a means by which we are able to appreciate the feelings of make initial tentative moves away from the mother. These others and to relate to others empathically. In other words, moves are increased during the next phase ofpracticing, as the affective domain provides an important component of the infant's burgeoning motor abilities and increased interest the continuous construction model. Whether affect deserves in the object world lead to longer periods of time playing the central or merely an important place in memory and the away from the mother. With cognitive advances in the middle subjective sense of continuity is controversial (Stern. 1988). of the second year, infants' awareness of separateness grows. but few would disgree that one of the most compelling con­ This dawning awareness of separateness and loss of omnipo­ tinuities posited is an individual 's sense of self. tence that characterize the rapprochement crisis are accom­ panied by an affective shift in the infant from joyous intoxi­ Development of the Sense of Self cation to increased petulance and sadness. Infant ambivalence about fusion and separateness during this period is inferred Development of the sense of self is fundamental to a from the well-described behavioral patterns of "shadowing" discussion of infant development because it is in the individ­ the mother and alternately "darting away" from her. Follow­ ual's identity that the domains of temperament, motivational ing resolution of the rapprochement crisis, infants begin to systems, competence, and affect become integrated. There is accept their separateness in the phase designated on the way an increasing consensus that organizational features of the to object constancy. sense of self give coherence and continuity to individual Despite the richness of Mahler's descriptions, the lack of experience (Damon and Hart , 1986). Although Horney methodological rigor has left unanswered the question of (1951), Kohut (1977), Kernberg (1975), and other psychoan­ whether the observations merely confirmed investigator biases alytic theorists have described the sense of self from the adult or critically tested specific hypotheses (Minde, 1981, 1982). perspective, infancy research has fostered new and important The theory of separation-individuation also appears both to 662 ZEANAH ET AL. overestimate and to underestimate the capacities of infants in new capacities for sharing attention, intentions, and affective the first 2 years of life (Horner, 1985; Stern, 1985a; Horner, states with an other (all of which have ample empirical 1988). A wealth ofresearch has established the infant's active support) ushers in the subjective sense of self. With this engagement with the world immediately after birth, which is addition to the emergent sense of self and the core sense of incompatible with Mahler's connotations of "autistic" disin­ self, the infant experiences a dramatic advance in relatedness. terest in, or avoidance of, human contact (Peterfreund, 1978; For the first time, the infant is aware of his or her ability to Lichtenberg, 1983; Horner, 1985; Stern, 1985a). Mahler's match and to mismatch mental states with an other. description of symbiosis too readily ascribes to infants capac­ The verbal sense ofselfemerges after the onset of the final ities for symbolic functioning (i.e., those necessary for hallu­ biobehavioral shift in the first 2 years. This new sense of self cinatory wish-fulfillment), that are well beyond anything that involves the ability to reflect on oneself and to use language has been demonstrated in infants at this age and also ignores to communicate about oneself to others. The advantages of what is known about capacities for differentiation (Horner, the verbal sense of self in vastly expanding the infant's expe­ 1985; Stern, 1985a). Feeling omnipotent is also incompatible riences and relatedness are obvious, but Stern also points out with infants' everyday perceptual experiences (Horner, 1985) a darker side of this new sense of self. The infant experiences and requires sophisticated reality-distorting capacities in the the limitations of language in rendering experiences, particu­ subjective experience of very young infants. Contrary evi­ larly those experienced in the domains of the other three dence suggests that in infancy the reality principle precedes sensesof self. "Language forces a space between interpersonal the pleasure principle, meaning that defensive operations, in experience as lived and as represented. And it is exactly across the psychodynamic sense, require cognitive capacities beyond this space that the connections and associations that constitute those available to 4-month-old infants (Stern, 1985a). neurotic behavior may form" (Stern, 1985a, p. 182). Finally, because of its derivation from the psychopathology In the third or fourth year of life, the infant begins to use of older psychotic children, the theory of separation-indivi­ language for more than objectifyingand labeling. For the first duation is fundamentally pathomorphic. The attribution to time, the infant demonstrates an ability to narrate his or her normal infants of characteristics of later psychopathology fits own life story. This momentous achievement marks the be­ with the fixation-regression model of psychopathology, but is ginning of an ability to change how one views oneself that is highly problematic in explaining normal development essential for psychodynamic psychotherapy. The narrative (Bowlby, 1969/1982; Peterfreund, 1978;Klein, 1980;Horner, sense of self describes the new domain of experience that 1985; Stern, 1985a). constructs a story from a variety of elements (e.g., actor, In striking contrast to Mahler's theory, Stern's (I985a) action, intentions, instrumentality, and context) drawn from theory of self development suggeststhat infants never subjec­ other senses of self (Stern, 1989a, b). The narrative self re­ tively experience undifferentiation; in fact, in his view they quires a qualitatively different mode of thought from problem are not capable of this since "only an observer who has enough solving or other kinds of talking. Although it is not yet clear perspective to know the future course of things can even why or how children begin to construct an autobiographical imagine an undifferentiated state" (p. 46). Instead, Stern history, it is clear that this history becomes the life story that envisions the infant as experiencing five relatively distinct an adult patient initially presents to a therapist. senses of self, each of which emerges in conjunction with the Unlike most other developmental theories that involve new capacities that accompany the biobehavioral shifts de­ sequential phases or stages,Stern's suggeststhat, once formed, scribed earlier. the senses of self operate continuously and simultaneously. Stern has proposed that in the first 2 months of life, infants None of them is necessarily dominant at any time in devel­ repeatedly experience the process of relating diverse objects opment nor is anyone necessarily linked to a particular and experiences in order to identify their invariant properties. clinical issue. Furthermore, subjective experience throughout Research indicates that some of this capacity for relating is development is organized not by phase-specificclinical issues innate, as infants' apparently possess a capacity for amodel such as orality, attachment, or mastery, but instead by the representation that allowsthem to "match" stimulation across emergent, core, subjective, verbal, and narrative sensesof self. different sensory modalities. Other means of relating are Regression is unnecessary in this conceptualization, because quickly learned, as networks of perceptions become organized of the continuing presence and growth of each of the five and integrated. Stern designates this sense of experiencing the senses of self throughout the lifespan. This is precisely the coming into being oforganization, the sense of emergent self. continuity implied by the continuous construction model. Following the first biobehavioral shift at 2 to 3 months, These two major theories of self development agree about another sense of self is added to the first. The core sense of the importance of the individual's subjectivity and the cen­ self is the physical self that is experienced as a coherent, trality of the sense of self in any attempt to understand willful, bounded physical entity with a unique affective life development. Both conceptualizations address the fundamen­ and history belonging to it. Stern cites a wealth of empirical tal human problem of being with others and being alone, and support for the infant's capacity for self-agency, self-coher­ they view this problem as a central organizer of human ence, self-affectivity,and self-historynecessary for a core sense experience. They also share much agreement about salient of self. The infant quickly learns that regulation of his or her clinical issues such as autonomy and what it feels like to be emotional states is dependent upon others (adult caregivers), with an other. Nevertheless, the nature of the individual's but even self-with-other experiences still belong entirely to subjectivity is radically different in the two systems. In addi­ the infant's own self. tion to vastly different conceptualizations of individual ex­ After the second and third biobehavioral shifts, the infant's perience, the theories of Mahler and Stern also understand PSYCHODYNAMICS: INFANT DEVELOPMENT RESEARCH 663 the origins of psychopathology differently. Mahler's work 1987).This assumes a constructive, narrative viewof memory illustrates the fixation-regression model of psychopathology, and recalI, consistent with recent research in cognitive psy­ with later problems understood as repetitions of infantile chology (Reiser et al., 1985). traumas. Further, the form of later pathology is determined by the sensitive period of self development in which the Relationship Patterns in Infants trauma occurred. Stem's conceptualization, on the other The cross-model perceptual capacity in young infants de­ hand, illustrates the continuous construction model, in which scribed earlier implies that a primitive representational capac­ there is no regression but only ingression into an ongoing ity exists even at birth. Other evidence supports the notion sense of self that is appropriate to a particular clinical issue. that infants "internalize," or retain patterns of experience in The form of later pathology is not necessarily related to the the first few months of life. During one phase of face-to-face developmental period in which it occurred originally. Since interactions with their 3-month-old infants, mothers were the self emerges in the context of the parent-infant relation­ instructed to "look depressed" (Cohn and Tronick, 1983). ship, a fundamental issue for both models is the degree to Infants became disorganized and distressed in response and, which early relationships experiences become internalized interestingly, they continued to exhibit distressed behavior for within the self during development. a period of time after their mothers resumed normal interac­ tion with them. Further, when infants of clinically depressed Internal Representations and Relationship Patterns mothers were compared to infants of nondepressed mothers Another domain in which continuity has been recognized in the same interactional paradigm, the infants of nonde­ amidst change is in patterns of relationships between infants pressed mothers became significantly more distressed during and their caregivers (Sroufe and Fleeson, 1986). A central the "look depressed" phase of interaction (Field, 1984). This tenet of virtualIy all psychodynam ic theories has been the suggests that the depressed appearance of the nondepressed notion that adults recreate early relationship experiences in mothers was a greater violation of expectation than the de­ subsequent relationships. The "compulsion to repeat" was pressed appearance of the c1inicalIy depressed mothers. originalIy described by Freud (1920) in the context of neurotic Other empirical support relevant to the construct of internal repetitions of maladaptive behaviors. The degree to which the representations and the intergenerational transmission of re­ childhood experiences of an individual are important influ­ lationships has come from investigations of attachment in ences on that individual's subsequent parenting behavior, and infants and adults (Sroufe and Fleeson, 1986).Inspired by the the factors that increase or decrease the likelihood of repeti­ ethologic-attachment theory of Bowlby (1969/1982, 1973. tion, are of enormous clinical relevance. Relationship conti­ 1980). Ainsworth et al. (1978) developed the most widely nuity and repetition imply some capacity in infants to inter­ used assessment of attachment relationships, a laboratory nalize and carry forward relationship patterns. Recent re­ paradigm known as the Strange Situation. This procedure search has attempted to operationalize the construct of involves a series of increasingly stressful episodes for 11- to internal representations of relationship experiences in order 20-month-old infants. On the basis of the organization (rather to explain how infants experience and reenact relationship than the content) of the infant's reunion behavior with respect patterns. to the caregiver following a brief separation, it is possible to Internal representations are essentially memory structures classifythe infant's pattern of attachment to that caregiver. that re-present a version of lived experience to an individual. According to attachment theory, infants should use their Stem (1985a) has outlined the process by which they are attachment figures as a secure base to explore the novel formed. The infant lives a particular experience, and, when it environment provided by the Strange Situation. Following is over, the experience is instantaneously transformed into a the caregiver's return, an infant should seek to reestablish memory. After the infant has lived a number of similar interaction with the caregiver and, if distressed, should seek experiences, each of which has been transformed into a par­ comfort from the caregiver. When infants behave in this ticular memory, the infant abstracts an average version of the manner, they are classified as securely attached to that care­ experience. This abstract average of related memories of giver. In contrast, some infants seem surprisingly undistressed experiences is an internal representat ion. This is analogous to by their caregivers' leavingand actively avoid them on return , the development of natural categories of objects in memory ignoring the caregiver's bid for interaction and attending that has been described by Rosch et al. (1976) and of gener­ instead to toys. The relationship between these infants and alized event representations described by Nelson and Gruen­ caregiversis termed avoidant. Other infants protest vigorously del (1981). when their caregiver leaves the room during the Strange There is an increasing consensus that internal representa­ Situation procedure, but they behave ambivalently when their tions are organized hierachically from smalI units reflecting caregivers return, alternately demanding contact and then subjective experience to increasingly large networks reflecting resisting it. Relationships characterized by this pattern are more global appraisals (Stem, 1985a, 1989a, b; Bretherton, termed ambivalent. 1985, 1987).These large networks, termed "working models" What is being classified is not simply a temperamental trait by Bowlby(1969/1982), not only re-present lived experience, of the infant, as some have suggested (Chess and Thomas, but also are presumed to perceive and interpret incoming 1982; Chess, 1984; Kagan et al., 1987). If infants are assessed information selectively,to generate anticipations, and to guide in the Strange Situation with different caregivers,they behave behavior in relationships. They are not merely passive filters differently depending on the quality of attachment to the of experience but contribute towards an individual's active caregiver with whom they are assessed (Main and Weston, recreations of relationship experiences (Zeanah and Anders, 1981 ; Belsky et al., 1984; Grossman et aI., 1985). Independ- 664 ZEANAH ET AL. ence of Strange Situation classifications of the same infant attachment relationships. The Adult Attachment Interview with different caregivers emphasize that attachment classifi­ inquires about early relationships, separations, losses, and cations are relationship specific. Obviously, this does not other attachment-relevant experiences. Scoring relies not on preclude the possibility that temperamental traits exert indi­ the content of the adult's descriptions, but rather on the rect effects on attachment classification, through effects on organization ofthoughts and feelings, and qualitative aspects caregivers' sensitivity and responsiveness, for example. Never­ of descriptions. Thus, whether an adult describes early expe­ theless, Strange Situation classification apparently reflects the riences and relationships as good or bad is less important than infant's view ofhis or her relationship with a specific caregiver the degree to which the adult has integrated these experiences, at a particular point in time (Hinde, 1982). as reflected in having access to memories and feelings regard­ Strange Situation classifications are preceded by character­ ing significant attachment figures and experiences and coher­ istic patterns ofcaregiver interactive behavior at home during ently describing these early events and experiences. the preceding year. The most consistent finding is that care­ Following the attachment classification system of the givers who are sensitive and responsive (i. e., consistently Strange Situation for infants, adults are classifiedas dismissing nurturant, attentive, and nonintrusive while interacting with ofattachment (corresponds to avoidance in infants), autono­ their infants) have secure attachment relationships with them mous with respect to attachment (corresponds to secure in at I year (Ainsworth et al., 1978; Egeland and Farber, 1984; infants), or preoccupied by past attachments (corresponds to Belsky et al., 1984; Grossman et al., 1985). Because Strange ambivalence in infants). Adults classified as autonomous Situation classifications are preceded by certain interactive value relationships, maintain a balanced view oftheir role in patterns at home, they may be understood as "summary relationships and a tolerance for imperfection in themselves outcome scores" of infants' relationship experiences with and others and are coherent in describing early experiences. primary caregivers in the first year of life. The working Adults classified as dismissing or preoccupied lack these qual­ hypothesis ofattachment theory is that these summary scores ities. The dismissing group describes attachment relationships reflect the infant's internal representation of the relationship as unimportant or claim to be unaffected by them and often with a particular caregiver. have difficulty remembering early experiences. The specific Infant attachment classifications derived from behavior memories they do recall often contradict their idealized global during the Strange Situation procedure are apparently stable. descriptions of experiences. The preoccupied group still seems Under conditions of low stress and adequate support, the dependent upon and overly concerned with their families and infant's classification with the caregiver remains stable be­ early experiences and are often still struggling to please family tween 12 and 18 months (Waters, 1978; Main and Weston, members. This group generally exhibits considerable difficulty 1981; Belsky et al., 1984). When attachment classification providing a coherent depiction ofrelationships or experiences. changes from secure to insecure, it is related to an increase in stresses or a decrease in supports (Vaughn et al., 1979; Intergenerational Continuity Thompson et aI., 1982). In other words, factors that would If the Strange Situation measures the "infant side" of the be expected to disrupt caregiver-infant relationships and there­ attachment relationship and the Adult Attachment Interview fore to change infant representations of those relationships measures the "adult side" ofthe attachment relationship, then have been demonstrated to affect them in predictable ways. psychodynamic assertions about intergenerational transmis­ In support of psychodynamic assertions about carrying sion lead us to expect convergence in the measures. Indeed, forward relationship patterns, attachment classification of as Main and Goldwyn (1984, 1989) have pointed out, there infants predicts subsequent psychosocial adaptation (Sroufe, are striking parallels in the patterns of infant behavior in the 1988). Infants who are securely attached to their primary Strange Situation procedure and the patterns ofadult patterns caregivers at 12 months of age are more autonomous at age of language, thought, and memory in the interview. For 2 years (Matas et al., 1978), have more advanced symbolic instance, adults classified as dismissing tum their attention play development (Slade, 1987), are more socially competent away from attachment relevant information during the Adult with peers and more ego-resilient (flexible, self-reliant, curi­ Attachment Interview, while simultaneously asserting their ous, involved) in preschool and kindergarten years (Arend et independence. They emphasize the normalcy of their experi­ al., 1979; Waters et aI., 1979; Troy and Sroufe, 1987; Oppen­ ences and minimize adversity or the effects of adversity. In heim et aI., 1988). School-aged children previously classified similar fashion, infants classified as avoidant tum their atten­ as securely attached demonstrate more competent overall tion away from their caregiver in the Strange Situation, as if functioning (Main et al., 1985), higher self-esteem (Cassidy, they are dismissing the importance of the parent, their rela­ 1988), and less psychopathology (Lewis et al., 1984; Erickson tionship to the parent, and their own need for comfort. Infants et aI., 1985) than those previously classified as insecurely classified ambivalent are either angrily inconsolable or pas­ attached. sively ineffective during the reunion episodes of the Strange Situation, and adults classified preoccupied are angrily en­ InternalRepresentations in Adults gaged in an unsuccessful struggle to please their parents or Believing that individual differences in internal represen­ passively incoherent about poorly defined childhood experi­ tations ofattachment ought to be measurable in adults as well ences. as infants, Main and her colleagues provided a second major Recent investigations with the Adult Attachment Interview conceptual and methodological advance in attachment re­ have examined the agreement between adults' classifications search. George et al. (unpublished manuscript) developed a in the Adult Attachment Interview and their children's clas­ measure to classify adults' description oftheir own childhood sification to that adult in the Strange Situation. In a prelimi- PSYCHODYNAMICS: INFANT DEVELOPMENT RESEARCH 665 nary investigation, Main et al. (1985) found a 76% agreement merely whether biological propensities or environmental between mothers' AAI classification and their children's stressors lead to maladaptive behavior, but, also, are the Strange Situation classification measured 5 years previously. disturbances fixed, or can they be altered by therapy or by Eichberg (1987), who interviewed mothers 6 to 12 months other forms of treatment? Emde (1988b) has called the dis­ after their infant's were observed in the Strange Situation, crepancy between the fixed rigidity of neurotic character found an 82% exact agreement between mothers' attachment pathology on the one hand, and the paucity of demonstrable classifications and the classification of their infants' relation­ behavioral continuities in individual infants on the other, the ships to them. Results from these investigations and the central developmental paradox. The key to understanding parallels in infant and adult behavior described above imply this paradox is in the recognition that continuities are at the that the caregivers and infants share similar patterns of proc­ level of subjective experience and relationship patterns rather essing attachment information and of affective arousal; in than individual behaviors. An infant who avoids his or her other words, similar internal representations of their relation­ mother after a brief separation at 12 months of age will be ship. These similarities in how caregivers and their young more dependent on the preschool teacher and more likely to children represent their relationship provide valuable prelim­ victimize classmates. This same infant is also at increased risk inary insights into the process by which relationship patterns for subsequent psychopathology (Sroufe, 1983; Erickson et may be carried forward across generations. al., 1985; Troy and Sroufe, 1987). Continuity in this case is Mothers' attachment classifications are overwhelmingly clearly not at the level of individual behaviors in the infant more insecure in clinical populations (Crowell and Feldman, but in the relationship pattern with the primary caregiver and in press; Benoit et aI., in press) and are related to their the internal insecurity that colors his or her subjective expe­ interactive behaviors with their young children (Crowell and rience. Feldman, in press). These findings provide preliminary evi­ A related question with origins in early psychoanalytic dence that internal representations of attachment in adults theory asks whether single traumatic early experiences inex­ may be meaningful indices of how they relate as caregivers to orably alter subsequent behavior. There is increasing evidence their children. In terms of the continuous construction model, from the research literature that massive psychological and/ internal representations occupy a pivotal position between or physical trauma experienced by individuals at any age have internal subjective experience and outward interactional be­ long-lasting effects on behavior (Eth and Pynoos, 1985). havior. As such, they become a major focus of treatment. Nevertheless, especially in early infancy, single traumatic events have limited effects. More important are the ongoing Discussion contexts in which these traumatic events occur. Early, brief The foregoing review highlights significant areas of research separations from caregiversand abrupt weanings, for instance, demonstrating that infants in the first 3 years of life are well are considered far less traumatic for development than dis­ organized to respond to, interact with, and actively engage ordered relationships involving the infant in a pattern of their environment. These capacities develop in orderly pat­ insensitive caregiving. Context provides the arena in which terns with discrete developmental shifts in levels of function environmental stressesinteract with individual biological pro­ and adaptation that change both the infant's experience of pensities to shape individual personalities with unique vul­ the environment and the environment's experience of the nerabilities and invulnerabilities. infant. Discontinuity in individual behaviors occurs within As described in this review,current research in infancy also the subjective continuity of the infant's developing sense of supports a paradigm shift away from the fixation-regression self and the stability of early relationship patterns. But what model and toward the continuous construction model of are the implications of these findings for current theory and development and psychopathology. Various clinical issues practice? may originate at any point in development and exert their The transactional model of development has provided an influence on self experience or relationship experiences. Em­ understanding of psychopathology that takes us beyond the phasis is on the individual's ongoing dynamic patterns of linear causality of nature-nurture dualism. Current research internal representations and interpersonal relationships. This in infancy has substantiated the view that reductionistic ap­ conceptualization has important implications for treatment, proaches to understanding behavior are not likely to be suc­ as well. cessful. Development requires both nature and nurture, and If a patient's major conflictual themes are no longer tied to it is the context in which these experiences interact that specific phases of development but instead are derived from require understanding in order to predict subsequent psycho­ the reconstructed "moments" that emerge in the patient's pathology. Even with continuities in developmental processes reported life story, then psychotherapy with adults must be­ that are obvious, such as the social deficits in autistic children, come more individualized (Stem, 1985a). Preconceived for­ it is the fit between continuously present markers, biological mulations concerning ontogenetic origins of psychopathology and environmental, that contributes substantially to outcome. are no longer valid. For example, control conflicts no longer Contexts are especiallyimportant in infancy, when psychiatric imply inevitable anal trauma. Furthermore, recovering the disorders are less clearly localized within an individual and putative origins of clinical trauma is useful only to the extent more appropriately within an infant's specific important re­ that it enables the individual to make changes in the here and lationships (Anders, 1989). These relationships provide the now, to challenge representational distortions, and to gain initial organizing context in which the fit between individual some conscious control over current behavior. The link with subjective experience and environmental demands occurs. the past may be most useful in helping the individual appre­ From the psychodynamic perspective, the question is not ciate the model that guides his or her current behavior and to 666 ZEANAH ET AL. evaluate more consciously its appropriateness in the present, J. D. Call, E. Galenson, and R. L. Tyson, New York: BasicBooks. Arend, R., Grove, F. L., & Sroufe, L. A. (1979), Continuity of especially as regards intimate relationships and providing for individual adaptation from infancy to kindergarten. Child Dev., the young. In treating children, the focus is less on retracing 50:950-959. origins and meanings of specific symptomatic behaviors and Barrett, K. C. & Campos, J. J. (1987), Perspectives on emotional more on changing current environmental contexts that sup­ development II. In: Handbook of Infant Development, ed. J. D. port the maladaptive behavior or symptomatology. Osofsky. New York: John Wiley& Sons. Basch, M. (1988), Understanding Psychotherapy. New York: Basic Since the crucible for early experiences is actually the Books. context ofprimary relationships, the construct ofrelationship Bates, J. E. (1980), The concept of difficult temperament. Merrill­ psychopathology has become the focus of therapeutic atten­ Palmer Quarterly, 26:299-319. tion. The transference relationship both with the therapist Bates,J. E. (1983), Issuesin the assessmentof difficulttemperament. and with significant others in the individual's life continues Merrill-Palmer Quarterly, 26:299-319. Belsky, J., Rovine, M., & Taylor, D. (1984),The Pennsylvaniainfant to be the model for this process. The characteristics that and family development project III. Child Dev., 55:718-728. underlie the primary attachment relationship, namely emo­ Benoit, D., Zeanah, C. H., & Barton, M. L. (1989), Maternal attach­ tional availability, dependability, empathic attunement, sen­ ment disturbancesin failureto thrive. Infant Mental Health Journal sitivity to developmental needs, and provision ofcomfort and (in press). Bowlby, J. (1969/1982), Attachment. New York: BasicBooks. security, are also prerequisites of the therapeutic relationship Bowlby, J. (1973), Separation. New York: BasicBooks. (Peterfreund, 1983). The therapist's presence, availability, and Bowlby, J. (1980), Loss. New York: BasicBooks. sensitivity provide opportunities for intense relationship ex­ Bretherton, I. (1985),Attachment theory: retrospectand prospect. In: periences that provide a context for understanding the indi­ Growing Points in Attachment Theory and Research, ed. I. Breth­ vidual's current internal representation of intimate relation­ erton and E. Waters. SRCD Monographs, Vol. 49, No.6, Serial no. 209. ships. Bretherton, I. (1987). New perspectiveson attachment relations. In: The complexities of human development are mirrored in Handbook ofInfant Development, ed. J. D. Osofsky. New York: the complexities of psychotherapy. Just as there is no single Wiley. determinant of psychopathological behavior, there is also no Bretheron, I. & Beeghley-Srnith, M. (1982), Talking about internal states. Developmental Psychology, 18:906-921. single therapy that can resolve every disorder. Multiple ther­ Brunner, J. S. (197'1), Early social interaction and language acquisi­ apeutic approaches are frequently necessary, depending upon tion. In: Studies in Mother-Infant Interaction, ed. H. R. Schaffer. the particular psychopathology. Infancy research, however, London: Academic Press. supports the psychodynamic assertion that amelioration of Buss, A. H. & Plomin, R. (1984), Temperament: Early Developing relationship pathology requires therapeutic relationship ex­ Personality Traits. Hillsdale, N.J.: Lawrence Erlbaum. Cameron, J. R. (1978), Parental treatment, children's temperament, periences, with particular strengths implicit in the transference and the risk of childhood behavioral problems 2. Amer. J. Orthop­ relationship as a major potential vehicle for change. sychiat.48:140-147. Whether the psychotherapeutic process can foster signifi­ Campos, J. J. & Stenburg, C. (1981), Perception, appraisal and cant changes in those aspects of personality and psychopath­ emotion. In: Infant Social Cognition, ed. M. Lamb & L. Sherrod. Hillsdale, N.J.: Erlbaum. ology that derive from internal representations, or whether Carey, W. (1986), Interactions of temperament in clinicalconditions. psychotherapy merely provides understanding of individual Advances in Developmental and Behavioral Pediatrics, 6:83-115. patterns of vulnerability and invulnerability remains contro­ Cassidy, J. (1988), Child-mother attachment and the self in six-year­ versial. Psychotherapy with children and families have dem­ olds. Child Dev. 59:121-134. onstrated how difficult it is, even at an early age, to induce Chess,S. (1984), What does behavior in the Strange Situation tell us? Behavioral and Brain Sciences, 7:148-149. significant characterologic changes. Even removing young Chess, S. & Thomas, A. E. (1982), Infant bonding: mystique and children from extremely pathological environments and plac­ reality.Amer. J. Orthopsychiat., 52:213-222. ing them in more optimal settings does not provide magical Clarke, A. M. & Clarke, A. D. B.(1976), Early Experience: Myth and transformations of personality or behavior. Simplistic thera­ Evidence. London: Open Books. Cohn, J. F. & Tronick, E. Z. (1983), Three month old infants' peutic interventions are as limited as unitary etiologic for­ reactions to simulated maternal depression. Child Dev., 54:185­ mulations. Complex, multidetermined problems in develop­ 193. ment require complex, multimodal solutions. The contribu­ Crowell, J. & Feldman, S. (1988), The effects of mothers' internal tion of psychodynamic psychotherapy seems best suited to models of relationships and children's behavioral and develop­ recreating the context of early development in order to reset mental status on mother-child interaction. Child Dev., 59:1273­ 1285. the level of vulnerability and invulnerability for disordered Damon, W. & Hart, D. (1986), Stability and change in children's behavior. self-understanding. Social Cognition, 4:102-118. Egeland, B. & Farber, E. A. (1984l, Infant-mother attachment. Child Dev., 55:753-771. References Eichberg, C. G. (1987), Quality of infant-parent attachment. Paper Achenbach,T. M., McConaugh,S. H., & Howell,C. T. (1987),Child/ presented to the Biennial Meeting of the Society for Reserach in adolescent behavioral and emotional problems. Psychol. Bull., Child Development. 101:213-232. Ekman, P., Friesen, W. V., & Ellsworth, P. (1972), Emotion in the Ainsworth, M. D. S., Blehar, M., Waters, E., & Wall, S. (1978), Human Face. New York: Pergammom Press. Patterns ofAttachment. Hillsdale, N.J.: LawrenceErlbaum. Emde, R. N. (1985), The affective self: continuities and transforma­ Anders, T. F. (1989), Toward a nosology of relationship disorders. tions from infancy. In: Frontiers ofInfant Psychiatry, ed. J. Call, In: Relationships and Relationship Disorders, ed. A. J. Sameroff E. Galenson, and R. Stimston. Vol. 2, New York: BasicBooks. and R. N. Emde, New York: BasicBooks. Emde, R. N. (l988a), Development terminable and interminable I. Anders,T. F. & Zeanah, C. H. (1984),Early infant development from Int. J. Psychoanal., 69:23-42. a biological point of view. In: Frontiers ofInfant Psychiatry II, ed. Emde, R. N. (1988b), Development terminable and interminable II. PSYCHODYNAMICS: INFANT DEVELOPMENT RESEARCH 667

Int . J. Psychoanal., 69:283-296. In: Continuities and Discontinuities in Development, ed. R. N. Emde, R. N., & Easterbrooks, M. A. (1985), Assessing emotional Emde and R. J. Harmon. New York: Plenum Press. availability in early development. In: Early Identification q{ Chil­ Kagan , J. (I984b), The Nature of the Child. New York: Basic Books. dren at Risk, ed. W. K. Frakenberg, R. N. Emde, & J. W. Sullivan. Kagan , J., Kearsley , R. B., & Zelazo, P. R. (1978), Infan cy: Its pl.ace New York: Plenum Press. in Human Development. Cambridge, Mass.: Harvard University Emde, R. N., Gaensbauer, T., & Harmon, R. J. (1976), Emotional Press. expression in infancy. Psychological Issues, Monograph 37. New Kagan, J., Reznick, J. S., & Snidman, N. (1987) , Temperamental York : International Universities Press. variation in response to the unfamiliar. In: Perinatal Development: Emde, R. N., & Harmon, R. J. (1972), Endogenous and exogenous APsychobiological Perspective, ed. N. A. Krasnegor, E. M. Blass, smiling systems in early infancy. J. Am. Acad. Child Psychiatry, M. A. Hofer, and W. P. Smotherman. Orlando, Fla.: Academic 11:177-200. Press. Emde, R. N. & Harmon, R. J. (1984), Entering a new era in the Kagan, J., Reznick, J. S., & Snidman, N. (1988), Biological bases of search for developmental continuities. In: Continuities and Discon­ childhood shyness. Science, 240:167-171. tinuities in Development, ed. R. N. Emde and R. J. Harmon. New Kaye, K. (1982), The Mental Life ofBabies. Chicago: University of York: Plenum Press. Chicago Press. Erickson, M. E, Sroufe, L. A., & Egeland, B. (1985), The relationship Kernberg, O. (1975), Borderline Conditions and Pathological Narcis­ between quality ofattachment and behavior problems in preschool sism. New York: Aronson. in a high risk sample. In: Growing Points in Attachment Theory Klein , M. (1980), On Mahler's autistic and symbiotic phase. Psycho­ and Research, ed. I. Bretherton and E. Waters. SRCD Monographs analysis and Contemporary Thought, 4:69-105. Vol. 49, No .6, Serial no. 209. Klinnert, M. D., Campos, J. J., Sorce, J. F., Emde, R. N., & Svejda, Eth, S. & Pynoos, R. (1985), Posttraumatic Stress Disorder in Chil­ M. (1982), Emotions as behavior regulators. In: Emotions in Early dren. Washington, D.C.: American Psychiatric Association Press. Development, ed. R. Plutchik and H. Kellerman. New York : Aca­ Field , T. (1984) , Early interactions between infants and their post­ demic Press. partum depressed mothers. Infant Behavior and Development, Klinnert, M. D., Sorce, J. F., Emde, R. N., Stenberg, c., & Gaens­ 7:527-532. bauer, T. (1984 ), Continuities and change in early emotional life. Field, T. & Greenburg, R. (1982), Temperarment ratings by parents, In: Continuities and Discontinuities in Development, R. N. Emde teachers of infants, toddlers and preschool children. Child Dev., and R. J. Harmon. New York : Plenum. 53:160-163. Kohut, H. (1977), Analysis ofthe Self New York : Aronson. Freud, S. (1920) , Beyond the Pleasure Principle. In: The Complete Lamb, M. E. & Sherrod, L. R. (1981) , Infant Social Cognition . Works, Vol. 18. London: Hogarth Press. Hillsdale, N.J.: Lawrence Erlbaum. Freud, S. (1940), An Outline of Psychoanalysis. In: The Complete Lewis, M. & Brooks-Gunn, J. (1979), Social Cognition and the Works, Vol. 23. London: Hogarth Press. Acquisition ofSelf New York : International Universities Press. Frodi, A., Bridges, L" & Grolnick, W. (1985), Correlates of mastery­ Lewis, M., Feiring, C; McGuffog, C.. & Jasher, J. (1984) , Pred icting related behavior. Child Dev., 56:1291-1298. psychopathology in six-year-olds from early social relation. Child Goldsmith, H. H., Buss, A. H., Plomin, R., Rothbart, M. K., Thomas, Dev., 55:123-126. A., Chess, S., Hinde, R. A., & McCall , R. B. ( 1987), Roundtable: Lichtenberg, J. D. (1983) , Psychoanalysis and Infant Research. Hills­ what is temperament? four approaches. Child Dev., 58:505-529. dale, N.J.: Analytic Press. Graham, P., Rutter, M. & George, S. (1973), Temperamental char­ Lipsitt, L. P. (1983), Advances in Infancy Research, Vol. 2. Norwood, acteristics as predictors of behavior disorders in children. Amer. J. N.J .: Ablex . Orthopsychiat., 43:328-339. Macturk, R. H., Hunter, F. T., McCarthy, M. E., Vietze, P. M., & Grolnick, W., Frodi, A., & Bridges, L. (1984), Maternal control style McQuiston, S. (1985), Social mastery motivation in Down syn­ and the mastery motivation ofone·year-olds.lnfant Mental Health drome and nondelayed infants. Topics in Early Childhood/Special Journal, 5:72-82. Education, 4:93-109. Grossman, K., Grossman, K. E.. Spangler, G., Suess, G., & Unzner, Macturk, R. H., McCarthy, M. E., Vietze, P. M., & Yarrow, L. J. L. (1985), Maternal sensitivity and newborns' orientation responses (1987), Sequential analysis of mastery behavior in 6 and 12 month as related to quality of attachment in northern Germany. In: old infants. Developmental Psychology, 23: 199-203. Growing Points in Attachment The ory and Research. ed. I. Breth­ Mahler, M., Pine, E, & Bergman, A. (1975), Psychological Birth of erton and E. Waters. SRCD Monographs, VOl. 49, No.6, Serial the Human Infant. New York: Basic Books. no . 209. Main, M. & Goldwyn, R. (1984) , Predicting a mother's rejection of Harter, S. (1978), Effectance motivation reconsidered. Human De­ her infant from representations of her own childhood experiences. velop., 21:34-64. International Journal ofChild Abuse and Neglect, 7:203-217. Hinde, R. A. (1982), Attachment: some conceptual and biological Main, M. & Goldwyn, R. (1989), Interview based adult attachment issues. In: the Place ofAttachment in Human Behavior, ed. C. M. classification. Developmental Psychology (in press). Parkes and J. Stevenson-Hinde. New York: Basic Books. Main, M., Kaplan, N., & Cassidy, J. (1985), Security of attachment Horner, T. M. (1985), The psychic life of the young infant. Amer. J in infancy, childhood, and adulthood. In: Growing Points in At· Orthopsyciat., 55:324-344. tachment Theon' and Research, ed. I. Bretherton & E. Waters, Horner, T. M. (1988) , Rapprochement in the psychic life of the SRCD Monographs, Vol. 49, No.6, Serial no . 209. toddler. Amer. J Orthopsychiat., 58:4-15. Main, M. & Weston, D. (1981), Security of attachment to mother Horney, K. (1951) , Neurosis andHuman Growth . New York: Norton. and to father. Child Dev., 52:932-940. Hubert, N. c.. Wachs, T . D., Peters-Martin, P., & Gandour, M. J. Malatesta, C. Z. (1981), Infant emotion and the vocal affect lexicon. (1982), The study of early temperament. Child Dev., 53:571-600. Motivation and Emotion, 5:1-23. Hunt, J. M. (1965), Intrinsic moti vation and its role in psychological Matas, L., Arend, R. A., & Sroufe, L. A. (1978), Continu ity of development. In: Nebra ska Symposium on Motivation, Vol. 13, adaption in the second year. Child Dev., 49:547-556. ed. D. Levine . Lincoln, Nebraska: University of Nebraska Press. McCall, R. B., Eichorn, D., & Hogarty, P. (1977) , Transition s in Izard, C. E. (1971), The Face of Emotions. New York: Appleton­ Early Mental Development, SRCD Monographs 42 (1177) . Century-Crofts. Mebert, C. J. & Kalinowsky, M. E (1986), Parents' expectations and Izard, C. E. & Malatesta, C. Z. (1987), Perspectives on Emotional perceptions of infant temperament. Infant Behavior and Develop­ Development I. New York: Wiley . ment, 9:321-334. Jennings, K. E., Connor, R. E., & Stegman, C. E. (1988), Does a Minde, K. K. (9181), Review of The Selected Papers of Margaret physical handicap alter the development of mastery motivation Mahler, New York: Jason Aronson. Journal ofthe American Acad­ during the preschool years? J. Am. Acad. Child Adolesc. Psychiatry, emy ofChild Psychiatry, 20:426-428. 27:312-317. Minde, K. K. (1982), Dr. Minde replies to Dr. Mahler. J. Am. Acad. Kagan, J. (1984a), Continuity and change in the opening years of life. Child Psychiatry, 21:94- 95. 668 ZEANAH ET AL.

Morgan, G. A. & Harmon, R. J. (1984), Developmental transforma­ Stern, D. N. (l989a), The representation of relational patterns. In: tions in mastery motivation. In: Continuities and Discontinuities Relationships and Relationship Disorders, ed. A. J. Sameroff and in Development, ed. R. N. Emde and R. J. Harmon. New York: R. N. Emde. New York: Basic Books. Plenum Press. Stern, D. N. (l989b), Crib narratives from a psychoanalytic perspec­ Nelson, K. & Gruendel, J. (1981), Generalized event representations. tive. In: Narratives From The Crib, ed. K. Nelson. Cambridge, In: Advances in Developmental Psychology, Vol. I, ed. M. E. Lamb Mass.: Harvard University Press. and A. L. Brown. Hillsdale, N.J.: Lawrence Erlbaum. Thomas, A., Chess, S., Birch, H. G., Hertzig, ME., & Korn, S. (1963), Oppenheim, D., Sagi, A., & Lamb, M. (1988), Infant-adult attach­ Behavioral Individuality in Early Childhood. New York: New York ments on the kibbutz and their relation to socioemotional devel­ University Press. opment 4 years later. Developmental Psychology, 24:427-433. Thomas, A., Chess, S., & Birch, H. G. (1968), Temperament and Peterfreund, E. (1978), Some critical comments on psychoanalytic Behavior Disorders in Children. New York: New York University conceptualizations of infancy. Int. J. Psychoanal., 59:427-441. Press. Peterfreund, E. (1983), The Process ofPsychoanalytic Therapy. Hills­ Thomas, A., Chess, S., & Korn, S. J. (1982), The reality of difficult dale, N.J.: Lawrence Erlbaum. temperament. Merrill-Palmer Quarterly, 28:1-20. Plomin, R. (1987), Developmental behavioral genetics and infancy. Thompson, R. A., Lamb, M. E., & Estes, D. (1982), Stability of In: Handbook ofInfant Development, ed. J. D. Osofsky. New York: infant-mother attachment and its relationship to changing life Wiley. circumstances in an unselected middle class sample. Child Dev., Reiser, B. J., Black, J. B., & Abelson, R. P. (1985), Knowledge 53:144-148. structures in the organization and retrieval of autobiographical Tronick, E. Z., Cohn, J. E., & Shea, E. (1986), The transfer ofaffect memories. Cognitive Psychology, 17:89-137. between mothers and infants. In: Affective Development in Infancy, Rosch, E., Mervis, C. B., Gray, W. D., Johnson, D. M., & Boyes­ ed. M. Yogman and T. B. Brazelton. Norwood, N.J.: Ablex, Braem, P. (1976), Basic objects in natural categories. Cognitive Troy, M. & Sroufe, L. A. (1987), Victimization among preschoolers. Psychology, 8:382-439. J. Am. Acad. Child Adolesc. Psychiatry, 26:166-172. Rosenbaum, J. F., Biederman, J., Gersten, M., Hirschfield, D. R., Vaughn, B. E., Bradley, C. F., Joffe, L. S., Seifer, R., & Barglow, P. Meminger, S. R., Herman, J. B., Kagan, J., Reznick, S., & Snid­ (1987), Maternal characteristics measured prenatally predict ratings man, N. (1988), Behavioral inhibition in children of parents with oftemperamental difficulty on the carey infant temperament Ques­ panic disorder and agoraphobia: Arch. Gen. Psychiatry, 45:463­ tionnaire. Developmental Psychology, 23:152-161. 470. Vaughn, B., Egeland, B., Sroufe, L. A., & Waters, E. (1979), Individ­ Rothbart, M. K., & Goldsmith, H. H. (1985), Three approaches to ual differences in infant-mother attachment at twelve and eighteen the study of infant temperament. Developmental Review, 5:237­ months: stability and changes in families under stress. Child Dev., 260. 50:971-975. Rutter, M. (1987), Continuities and discontinuities from infancy. In: Vaughn, B. E., Taraldson, B. J., Crichton, L., & Egeland, B. (1981), Handbook of Infant development. ed. J. D. Osofsky. New York: The assessment of infant temperament. Infant Behavior and De­ Wiley. velopment, 4:1-17. Sameroff, A. J. & Chandler, M. J. (1975), Reproductive risk and the Waters, E. (1978), The reliability and stability of individual differ­ cotinuum of caretaking casualty. In: A Review of Child Develop­ ences in infant-mother attachment. Child Dev., 49:483-494. ment Research, Vol. 4. Chicago: University ofChicago Press. Waters, E. & Sroufe, L. A. (1983), Social competence as a develop­ Sameroff, A. J., Seifer, R., & Elias, P. K. (1982). Sociocultural mental construct. Developmental Review, 3:79-97. variability in infant temperament ratings. Child Dev., 53:164-173. Waters, E., Wippman, J., & Sroufe, L. A. (1979), Attachment, positive Scarr, S. & Kidd, K. K. (1983), Developmental behavior genetics. In: affect, and competence in the peer group. Child Dev., 54:821-829. Handbook of Child Psychology, 4th ed., ed. P. H. Mussen. New White, R. W. (1959), Motivation reconsidered: the concept of com­ York: Wiley. petence. Psychological Review, 66:297-333. Seifer, R. & Sameroff, A. J. (1986), The concept, measurement, and Wolff, P. (1966), The Causes, Controls, and Organization ofBehavior interpretation of temperament in young children. In: Advances in in the Neonate, Monograph 17. New York: International Univer­ Developmental and Behavioral Pediatrics, 7:1-43. sities Press. -- Vaughn, B. E., Lefever, G. & Smith, C. (in press), Relationships Wolkind, S. N. & DeSalis, W. (1982), Infant temperament, maternal among mastery motivation and attachment within a general model mental slate and child behaviour problems. In: Temperamental of competence. In: Perspectives on Mastery Motivation in Infants Differences in Infants and Young Children (Ciba Foundation Sym­ and Children. ed. P. M. Vietze & R. H. MacTurk. posium 89). London: Pitman Books Ltd. Slade, A. (1987), Quality of attachment and early symbolic play Yarrow, L. J., Morgan, G. A., Jennings, K. D., Harmon, R. J., & Developmental Psychology, 23:78-85. Gaiter, J. L. (1982), Infants' persistence at tasks. Infant Behavior Sroufe, L. A. (1983), Individual patterns ofadaptation from infancy and Development, 5:131-141. to preschool. In: Development and Policy Concerning Children with Yarrow, L., McQuiston, S., Macturk, R. H., Klein, R. P., & Vietze, Special Needs. Minnesota Symposium on Child Psychology, Vol. P. M. (1983), Assessment of mastery motivation during the first 16, ed. M. Perlmutter. Hillsdale, N.J.: Lawrence Erlbaum. year of life. Developmental Psychology, 19:159-171. Sroufe, L. A. (1988), The role of infant-caregiver attachment in Yarrow, L. J., MacTurk, R. H., Vietze, P. M., McCarthy, M. E., development. In: Clinical Aspects ofAttachment, ed. J. Belsky and Klein, R. P., & McQuiston, S. (1984), Developmental course of T. Nesworski. Hillsdale, N.J.: Lawrence Erlbaum. parental stimulation and its relationship to mastery motivation Sroufe, L. A. & Fleeson, J. (1986), Attachment and the construction during infancy. Developmental Psychology, 20;492-503. of relationships. In: The Nature and Development ofRelationships, Zeanah, C. H. & Anders, T. F. (1987), Subjectivity in parent-infant ed. W. W. Hartup and Z. Rubin. Hillsdale, N.J.: Lawrence Erl­ relationships. Infant Mental Health Journal, 8:237-250. baum. Zeanah, C. H., Keener, M. A., & Anders, T. F. (l986a), Adolescent Sroufe, L. A. & Waters, E. (1977), Attachment as an organizational mothers' prenatal fantasies and working models of their infants, construct. Child Dev., 48: 1184-1199. Psychiatry, 49:193-203. Stem, D. N. (1977), The First Relationship: Infant and Mother. Zeanah, C. H., Keener, M. A., & Anders, T. F. (l986b), Developing Cambridge, Mass.: Harvard University Press. perceptions of temperament and their relation to mother and infant Stern, D. N. (l985a), The Interpersonal World of the Infant. New behavior. J. Child Psychol. Psychiatry, 27:499-512. York: Basic Books. Zeanah, C. H., Kenner, M. A., Anders, T. F., & Levine, R. (1986c), Stern, D. N. (1985b), Affect attunement. In: Frontiers of Infant Measuring difficult temperament in infancy, J. Dev. Behav. Pe­ Psychiatry Ll, ed. J. Call, E. Galenson, and R. L. Tyson. New York: diatr., 7:114-119. Basic Books. Zeanah, C. H., Keener, M. A., Stewart, L., & Anders, T. F. (1985), Stern, D. N. (1988), Affect in the context of the infant's lived Prenatal perception of infancy personality. J. Am. Acad. Child experience. Int. J. Psychoanal., 69:233-238. Psychiatry, 24:204-210.