Contemporary Psychoanalytic Perspectives on Attachment

Abstract

Bowlby's theoretical conception of attachment represented a highly original and substantively unique paradigm for understanding the psychological and social development of human infants, as well as the short- and longer-term sequelae of failures in attachment. His ideas have also been highly generative, leading to a vast scientific literature that now encompasses several different disciplinary domains. Moreover, unlike many psychoanalytic developmental concepts, Bowlby's attachment theory is regarded by many social scientists as eminently researchable. Nevertheless, despite increasing popularity, Bowlby's ideas about human development are not universally embraced by dynamically oriented clinicians, who may favor other theoretical frameworks to explain developmental phenomena or psychopathology, or perhaps both. How are we to understand Bowlby's ideas about attachment, separation, and loss in light of other psychoanalytic developmental theories? To address this issue, the “goodness of fit” between attachment theory and four important contemporary psychoanalytic frameworks for understanding human development will be examined. Selected because each offers a unique vision of development, the four are Harry Stack Sullivan's interpersonal model, Margaret Mahler's separation-individuation theory, Erik Erikson's epigenetic model, and Heinz Kohut's psychology of the self. Each will be examined in detail, with particular attention to controversies and points of convergence.

KEYWORDS:

• Bowlby
• attachment theory
• interpersonal psychiatry
• separation-individuation
• psychosocial epigenesis
• self-psychology

INTRODUCTION

Bowlby's theoretical conception of attachment represented a highly original and substantively unique paradigm for understanding the psychological and social development of human infants, as well as the short- and longer-term sequelae of failures in attachment. His ideas have also been highly generative, leading to a vast scientific literature that now encompasses several different disciplinary domains. Moreover, unlike many psychoanalytic developmental concepts, Bowlby's attachment theory is regarded by many social scientists as eminently researchable. His ideas also reflect the influence of his psychoanalytic predecessors and the British object relations tradition with which he identified. Bowlby's unique contribution appears to have resulted from the combined effect of several factors, foremost among them his systematic and empirical approach to the study of attachment, his willingness to challenge existing psychoanalytic ideas, and his ability to synthesize and integrate research and theoretical models from fields outside of psychoanalysis (most notably ethology). Nevertheless, despite increasing popularity, Bowlby's ideas about human development are not universally embraced by dynamically oriented clinicians, who may favor other theoretical frameworks to explain developmental phenomena or psychopathology, or perhaps both. How are we to understand Bowlby's ideas about attachment, separation, and loss in light of other psychoanalytic developmental theories?

To address this issue, the “goodness of fit” between attachment theory and four important contemporary psychoanalytic frameworks for understanding human development will be examined. Selected because each offers a unique vision of development, the four are Harry Stack Sullivan's interpersonal model, Margaret Mahler's separation-individuation theory, Erik Erikson's epigenetic model, and Heinz Kohut's psychology of the self. Each will be examined in detail, with particular attention to controversies and points of convergence.

SULLIVAN AND THE INTERPERSONAL MODEL

Harry Stack Sullivan (1892–1949), an American physician best known for his pioneering work in the psychiatric treatment of schizophrenic and other highly disturbed individuals, is generally credited with having founded the school of interpersonal psychoanalysis as well as having formulated its most important developmental and clinical ideas. While a medical student at the University of Chicago, Sullivan was greatly impressed by such important social scientists as sociologists George Herbert Mead and Robert Park and cultural anthropologist Edward Sapir; he was also influenced, though less directly, by John Dewey, and more arguably, by the social work pioneer and Hull House founder Jane Addams (Perry, 1964). Sullivan has been linked, more generally, with the American pragmatist school of philosophy; indeed, Sullivan's biographers have argued that his unique approach to clinical work reflected an enduring commitment to this intellectual tradition (Brandell & Perlman, 2010; Chapman, 1976; Mullahy, 1970; Perry, 1964, 1982).

Sullivan's influence during the 1940s and 1950s was considerable. Such important theorists as Frieda Fromm-Reichmann, Erich Fromm, and Clara Thompson, who, with Sullivan, are now considered architects of the interpersonal-relational tradition in psychoanalysis, were among those in his intellectual debt. However, Sullivan had never trained as a psychoanalyst, and despite the originality of his views regarding development and psychopathology, as well as a distinctive approach to psychotherapeutic treatment resulting from these ideas, his work remained somewhat outside the mainstream of American psychoanalytic thought and was virtually unknown elsewhere—that is, until the publication of Greenberg and Mitchell's epic volume on object relations theory in the early 1980s (Fonagy & Target, 2003; Greenberg & Mitchell, 1983). Since that time, Sullivan's ideas have achieved greater acceptance among psychoanalysts, though perhaps in part because some proponents of Sullivanian theory “have increasingly moved toward integrating Sullivan's thinking with contemporary systems of psychoanalytic thought rather than attempting to maintain his ideas in pure form” (Fonagy & Target, 2003, p. 205).

Sullivan's adherence to two theoretical positions, pragmatism and operationalism, as well as his early experience in treating schizophrenics, seem to have combined to create within him a profound dissatisfaction with traditional psychoanalytic developmental and clinical theories. Despite his great respect for Freud's creative genius, Sullivan expressed dismay over the failure of classical psychoanalysis to operationalize its theoretical precepts. Even such a fundamental psychoanalytic concept as the unconscious was, in Sullivan's view, problematic inasmuch as it could be neither observed nor measured and was therefore destined to remain wholly hypothetical. Sullivan also objected to Freudian theory on clinical grounds, believing that traditional psychoanalytic thinking had failed to adequately explain psychosis and other forms of severe psychopathology and, furthermore, that it did not offer a particularly useful approach to working clinically with such patients. At the very heart of Sullivan's critique of Freud was an elusively simple belief, one cited by psychoanalytic historians as the basis for designating Sullivan as a pioneering exponent of the “relational/structure” model: “The field of Psychiatry is the field of interpersonal relations—a personality can never be isolated from the complex of interpersonal relations in which the person lives and has his being” (Sullivan, 1940, p. 10).

Sullivan also endeavored to create an entirely new interpersonal “vocabulary,” intended to supplant the more abstract and less empirically anchored conceptual language of classical theory. Critics, however, have noted that the formulations of Sullivan's interpersonal theory are not significantly different epistemologically from those of other psychoanalytic schools (Singer, 1998); that the language of interpersonal theory was highly idiosyncratic and confusing, composed of terms that Sullivan himself had invented to describe various mental phenomena; that interpersonal psychoanalysis lacked a unified, integral theory; and even that interpersonal psychoanalysis was not especially original, merely representing “an attempt by Sullivan to recast the intrapsychic language of Freud into interpersonal terms” (Joseph Kepecs, MD, personal communication).

It is beyond the scope of this article to fully address such criticisms. Nor can we undertake a thorough discussion of Sullivan's clinical formulations. Rather, we shift our focus to Sullivan's model of human development, the most detailed exposition of which probably appears in The Interpersonal Theory of Psychiatry, a posthumously published edited collection of lectures given by Sullivan in the mid-1940s (Sullivan, 1953). As noted above, an overarching theme in Sullivan's work is that human infants are born into a relational milieu and that relational configurations evolve out of actual experience with others. Sullivan repeatedly stressed the assertion that human beings can be understood only within the “organism-environment complex” ¹ and are thus incapable of “definitive description in isolation” (Sullivan, 1930, p. 258). Accordingly, Sullivan's developmental model places supreme importance on the interpersonal field in tandem with the efforts that children—and later, adults—devise to maintain relatedness with significant others in their environments. Sullivan's developmental schema, formulated as a series of “heuristic stages,” consisted of infancy, childhood, the juvenile era, preadolescence, early adolescence, late adolescence, and adulthood or maturity; our focus will of necessity be limited to the earlier stages of development.

In infancy, Sullivan wrote, human beings possess innate physiological and emotional needs that require satisfaction, but their gratification depends upon the capacity of another individual, the mother, to experience the infant's tensions as her own and to act upon this experience (Fonagy & Target, 2003). Such complementary interactions are experienced as “tenderness” by the infant and establish a sort of interactional or interpersonal template for all subsequent experiences; ultimately, these serve to generate a need for tenderness in the infant. Sullivan believed that the infant's need for tenderness actually consists of a group of tensions for the most part derived from disequilibria involving the “physicochemical universe” existing both within and outside of the infant, and requiring interpersonal involvement. He also included one need that was not physicochemically induced, the need for contact, which reflected just what the term implies, a basic human need for “manipulations by and peripheral contact with” caregivers that is independent of the infant's other tensions (Sullivan, 1953, p. 40). ²

Sullivan theorized that infants begin to develop what he termed personifications early in their development. In the Sullivanian universe, a personification is defined as the representation of an infant's experience of others in the caretaking environment. Personifications often do not correspond closely to objective reality; they are most usefully thought of as “personal, idiosyncratic phenomenological representations” of self and others (Singer, 1998, p. 77). Thus, personifications are of two basic types—self and other—and acquire a predominantly positive (“good”) or negative (“bad”) valence depending on experiences of relative satisfaction or dissatisfaction. For example, the experience of nursing (at either bottle or breast), arguably a prototype for all subsequent interpersonal encounters in early development, initially results in the infant's personification of the nipple as either good (satisfying) or bad (frustrating). Gradually, via the repetition of many such experiences, and in consequence of the developing infant's capacity for perception and prehension, the nipple is integrated into a more expansive and multifaceted experience of the organism that provides the nipple—the infant's mother. When these experiences are gratifying, they lead to a personification of mother that is predominantly positive, and that may also include others (father, older siblings, and so forth) who, in addition to the mother, later become associated with the good feeding experience. Inversely, feeding experiences that evoke the infant's frustration and the human figures associated with such experiences are elaborated into a personification of the “bad” nipple/mother.

In this same example, the infant's personifications of the self are also closely connected to experiences with the “good” or “bad” nipple/mother. Sullivan assumed that infants are both highly egocentric and incapable of making discriminations between self and others in early development. ³ Consequently, the infant makes an assumption that she or he has performed some action or otherwise brought about the good mother personification; in effect, it is on account of something the infant “believes” she or he has done that evokes the good, satisfying nipple—an inner experience that Sullivan termed the good me. The same sort of association occurs when the nipple is non-gratifying or otherwise anxiety-generating—once again, it is something in the nature of the infant's behavior, a bad me, that has brought about this negative experience of the mother-infant dyad rather than independent actions taken by the mother herself (Singer, 1998).

Such primitive mental operations are further described by Sullivan as examples of parataxic thought, a primitive mode of cognition (much like the Freudian primary process) in which logical processes are largely absent and (oftentimes) erroneous connections are made on the basis of temporal contiguity or spatial sequencing rather than more formal reasoning using deductive or inductive processes. Frustrating or negative experiences with one caregiver resulting from that particular individual's limitations or unempathic responses to the infant not only may be incorrectly experienced as a shortcoming of the infant's but may in the future become generalized to other caregivers. Within a clinical context, it then becomes possible to understand the templates forged from early parataxic distortions as forming a basis for transference phenomena, though in a later, somewhat more evolved form (Singer, 1998).

In addition to “good” and “bad” self/other personifications, Sullivan also proposed a third variety, one associated with profoundly painful or terrifying interpersonal encounters. Such experiences, in Sullivan's view, might evoke qualitatively different representations of self and other, personifications that he called the “not me” and the “evil mother.” Although Sullivan believed “not me” and “evil mother” personifications to be universal, these representations are much less important or influential in normal or neurotic individuals than in those who suffer from schizophrenia or other severe psychopathology.

Interpersonal theory posits two kinds of motivation. The first, as intimated above, involves the gratification of bodily and emotional needs, including sexuality and intimacy. A second kind of motivation is associated with the experience of anxiety, as well as with those defensive maneuvers—which Sullivan called “security operations”—intended to alleviate such anxiety. Sullivan's premise is that both kinds of motivation need satisfaction and that the subject's strivings for security are comprehensible only within an interpersonal context. Indeed, in Sullivan's formulations, the intrapsychic emphasis of classical psychoanalytic theory, as well as Freud's later (1926) model of anxiety, have little utility or explanatory power. Whereas Freud understood anxiety relative to a hierarchy of danger situations (loss of the mother, loss of the mother's love, castration, and guilt), situations that may or may not be closely connected with interpersonal interactions, Sullivan emphasized mental distress, psychic pain, or upset activated in an interpersonal milieu. Furthermore, in Freud's final theoretical model, the function of anxiety as a signal of internally arising danger as well as the inextricable link between the drives and reality in the production of anxiety is underscored (Holzman, 1998). In Sullivan's theory, however, it is the caregiver's failure to manage or contain her or his anxiety or other distressing affects that educes anxiety in the infant. Perhaps foreshadowing ideas popularized by Heinz Kohut, Sullivan located the earliest experience of what might be called “empathy” in such experiences of “emotional contagion or communion” that “subtend the relationship of the infant with other significant” individuals (Sullivan, 1940, p. 17).

Sullivan's ideas are not truly compatible with those of Bowlby, although certain similarities are unmistakable. True, Sullivan's emphasis on the interpersonal matrix and his focus on the primacy of the infant's physiological needs seem more or less consonant with Bowlby's emphasis on attachment. Both theorists believed it impossible to understand human development on the basis of the libidinal-stage model; neither accepted the relatively scant role accorded to the environment and to the influence of caregivers on the infant's psychosocial development by traditional psychoanalysis; and neither accepted the rather limited role that classical formulations attributed to factors outside the infant in the production and/or management of anxiety. On closer examination, however, several significant differences emerge.

For example, Bowlby and Sullivan are not in agreement as to the importance of actual physical separation in the production of anxiety in the infant. Sullivan tended to locate anxiety, per his concept of empathic linkage, in the caregiver; in his view, it was ultimately a failure on the part of the care provider that led to the experience of anxiety in the infant. Although Sullivan believed that loneliness could be a devastating experience for adolescents and adults, he asserted that it was not an experience phenomenologically available to infants and young children (Sullivan, 1953). Indeed, in his critique of Sullivan, Bowlby suggests that Sullivan may never have fully grasped the importance of separation anxiety (Bowlby, 1973, p. 396). Bowlby was also critical of Sullivan's claim that “no action of the infant is consistently and frequently associated with the relief of anxiety” (Sullivan, 1953, p. 42), believing that Sullivan had overlooked the most obvious challenge to his assertion—the infant's relief when she or he is able to clutch the mother. Finally, Sullivan had maintained that self-appraisal in children is originally modeled after the way in which they are viewed by the significant adults in their lives, an idea that is firmly anchored in the Sullivanian theory of self/other personifications and that presupposes a more or less passive acceptance of such appraisals by the child. Bowlby's concept of internal working models, however, holds that children “not only passively accept” the views of others but also actively arrive at sometimes radically different appraisals of themselves or of others (Bowlby, 1980, pp. 234–235).

More-recent distillations of the Sullivanian theory, which has become rather intimately associated with the emerging relationalist perspective in contemporary psychoanalysis, highlight yet another, even more fundamental, difference between these two traditions. As Fonagy has observed,

relational theories, in part as a consequence of their historic origins, tend to repudiate the biological [emphasis added] in thinking about motivation and human nature. The adult human organism is not thought to be understandable in terms of other sorts of organisms, bestial or infantile, but has its own distinctive nature. It is not “driven” by “special” drives, but is the agent of many kinds of activities, all of which are devoted to the general project of creating, recreating, and expressing itself within its relational context. (2001, p. 131)

Attachment theory, of course, is either confirmed or refuted on the basis of both its biological foundations and its integration within a natural sciences context (Bowlby, 1981; Fonagy, 2001). Sullivanian interpersonal theory and its contemporary relationalist counterpart, however, are “more at home with postmodern deconstructive ideas” than they are with Bowlby's efforts to link brain and behavior and what Fonagy has referred to as the “reductionism of the biological context of attachment” (Fonagy, 2001, p. 32).

MAHLER’S THEORY OF SEPARATION-INDIVIDUATION

Margaret Mahler (1897–1985), who, like D. W. Winnicott and Melanie Klein, made a mid-career transition from pediatrics to psychoanalysis, was originally less focused on normal development than on arriving at a more satisfactory explanation for the development of psychotic disorders in children. Her early work with autistic and schizophrenic youngsters throughout the 1940s and 1950s, however, gradually led her to consider the importance of the role played by the human infant's “mediating partner,” the mother, not only within the matrix of psychopathology but also in the course of normal development. Mahler's theory is not without its critics, but her ideas regarding separation-individuation have been indisputably influential and are judged to occupy a crucial position in the history of psychoanalytic thought (Greenberg & Mitchell, 1983).

Although Mahler's theory of separation-individuation underwent numerous changes and revisions dating to its earliest distillations in the 1950s (e.g., Mahler, 1954; Mahler & Gosliner, 1955), ⁴ this discussion is, for the sake of clarity, concerned with the more fully evolved version that appeared in the mid-1970s (Mahler, Pine, & Bergman, 1975). By that time, Mahler had shifted her attention almost completely away from severe childhood psychopathology and was engaged in a project that sought to redefine the essential features of the normal infant's emergence from “formlessness to form”—emphasizing the relationship of child to mother rather than the classical focus on development of the libido.

This final version of Mahler's separation-individuation theory grew from a longitudinal research project involving mother-infant pairs conducted in a nursery setting (the Masters Children's Center in Manhattan). Characterizing infants as essentially nonrelated or objectless at birth (the normal autistic phase), Mahler and her colleagues described their gradual emergence via a period of maternal-infant symbiosis into four relatively discrete stages of separation and individuation: differentiation, practicing, rapprochement, and development of object constancy. Throughout this later work, Mahler characterized separation and individuation as complementary though nevertheless phenomenologically distinct processes. Separation was defined as the infant's gradual extrication from her or his symbiotic fusion with the mother, while individuation consisted of “those achievements marking the child's assumption of his own individual characteristics” (Greenberg & Mitchell, 1983; Mahler et al., 1975, p. 4).

The Normal Autistic Phase (Birth to 1 Month)

Mahler believed that in the first several weeks following birth, infants are relatively unresponsive to external stimuli and are largely incapable of making distinctions between inner and outer reality or, for that matter, between the self and the inanimate world. At this early stage of development, Mahler asserted, infants’ sleeplike states are disproportionately greater than their states of arousal, and their inner experience is a sort of primitive hallucinatory wish fulfillment in which the principal effort is directed toward the satisfaction of needs and, concomitantly, the achievement of physiological homeostasis (Mahler, 1979, p. 77). The most essential feature of Mahler's conception of normal autism lies in her depiction of the infant at this time as a closed psychological system, metaphorically encapsulated in a “quasi-solid stimulus barrier” or “autistic shell” that prevents penetration by external stimuli (Mahler & Furer, 1968), an idea underscored by Mahler's reference to Freud's (1911) analogy of the bird's egg as a model for autistic satisfaction of nutritional needs. Admittedly, this conception follows a classical psychoanalytic view of neonatal life, as Mahler seemed to anchor her description of this stage firmly in the tension-reduction language of libido theory (Freud, 1911). Infantile experience at this early stage of development is, from an object relations perspective, considered “objectless” (Greenberg & Mitchell, 1983).

The Phase of Normal Symbiosis (1–5 Months)

Symbiosis is a concept derived from biology to describe a close and mutually advantageous association between two organisms. Mindful of the fact that infantile dependence on caregivers is absolute, while that of mothers on infants is not, Mahler nevertheless chose to employ this term in a metaphorical way. To her, human symbiosis refers to a state of fusion between infant and mother in which self and not-self are not yet differentiated and in which there is only the faintest awareness of the distinction between inner and outer experience. Perhaps the most essential feature of human symbiosis, Mahler asserted, is the “hallucinatory or delusional, somatopsychic, omnipotent fusion with the representation of the mother and, in particular, [the] delusion of [a] common boundary” of two individuals who are in reality physically separate organisms (Mahler, 1979, p. 79). Infants during this developmental phase, however, are already capable of developmental achievements not possible just a few weeks before. There is a dim, though increasing, awareness of the mother as an object existing outside of the self, support for which is adduced in part by the social smile that infants first exhibit during this period of development. In addition, the infant now begins to invest libido toward the “periphery”—what Mahler referred to as the “dual unity” of mother and infant—rather than exclusively inward, as was the case during the first several weeks of life. Aided by the maturation of such autonomous ego functions as memory, infants begin to organize experience according to what is “good” (pleasurable) or “bad” (painful), and such experiences are subsequently stored as memory traces (Greenberg & Mitchell, 1983). It is during this phase, Mahler noted, that infants may be legitimately thought of in more psychological terms, although she nonetheless regarded them as “pre-objectal” during this phase (Mahler, 1979, p. 79).

The Differentiation Subphase (5–9 Months)

Sometime around four or five months, the first phase of the formal separation-individuation process is ushered in. Mahler used the term “hatching,” again borrowing from Freud's bird egg analogy, to describe the momentous psychological growth that occurs at this time. Although keenly interested in examining the mother's face and hair (what Mahler has termed “customs inspection”), differentiating infants are no longer totally dependent on their mothers’ bodies; they have begun to creep, crawl, and climb, and they also exhibit progress in the coordinated use of hands, mouth, and eyes (Mahler, 1979). There is continued growth and maturation of various ego functions, gradually permitting infants to make more reliable distinctions between that which belongs to inner experience and that which is perceived with the sense organs to exist outside the self, a critical developmental benchmark that makes possible increasingly clearer sensory discriminations between self and object. Peekaboo games offer the infant an opportunity to master some of the anxiety associated with the mother's comings and goings. At the same time, there is interest in the visual field beyond the mother, though each new person or object encountered is carefully, even soberly, compared to the mother, a process that Mahler referred to as “comparative scanning.” At six to eight months, these emerging skills and capabilities culminate in the infant's ability to clearly and consistently distinguish his or her mother from other figures in the environment, a developmental accomplishment signaled by the appearance of stranger anxiety.

The Practicing Subphase (9–15 Months)

This subphase, Mahler observed, is actually composed of two more or less distinct periods: early practicing and the practicing subphase proper. The early practicing period, at first coextensive with the latter part of the differentiation subphase, is defined by the infant's acquisition of the capacity for quadruped locomotion (crawling, climbing, paddling, and so forth). This important achievement makes possible more sustained forays into what Mahler termed “the intramural environment,” the world that exists beyond the mother. It is at this time, she believed, that infants’ belief in their own magical omnipotence reaches a peak and contributes to what Phyllis Greenacre called their “love affair with the world” (Greenberg & Mitchell, 1983). Now able to move independently, the infant finds a whole new universe to explore, one that is filled with novelty and excitement. At the same time, the infant periodically seeks the mother out as a “home base” for “emotional refueling,” further cementing the unique affectional bond that exists between the two of them. Yet another developmental task of early practicing is the “dramatic growth of the autonomous functions of the ego, a growth which occurs optimally in close proximity to the mother” (Greenberg & Mitchell, 1983, p. 276).

Mahler believed that “psychological birth” is coextensive with a monumental achievement—that of the capacity for upright locomotion. Feeling exhilaration both because of the seemingly limitless opportunities for exploration afforded by this new ability as well as by the avenue of “escape” it furnishes from the earlier symbiosis with the mother (Greenberg & Mitchell, 1983), the infant enters the practicing subphase proper. The mother is still not a fully separate person in the toddler's estimation, even now, and she continues to be treated as a “home base” for replenishment of the child's emotional needs. However, Mahler also noted that the mother must be willing to put aside her claims on her child's body at this juncture in the interest of promoting developmental progress; furthermore, she “must be willing to allow and even to enjoy [the child's] increasing capacity to operate at a distance from her” (Greenberg & Mitchell, 1983, p. 277).

The Rapprochement Subphase (15–24 Months)

The rapprochement subphase is initiated by the child's beginning awareness that the mother is in fact a separate person, that she operates independently of her toddler, and is therefore not always physically available to furnish assistance in the toddler's ongoing efforts to negotiate environmental challenges. Indeed, one of the hallmarks of the rapprochement subphase is the toddler's increasing—and at times painful—awareness that she or he is a rather small person in a big world. During the practicing period, the toddler's burgeoning sense of separateness was at first a cause for exhilaration, but now the earlier sense of omnipotence and the ideal self of which it was a concomitant have been replaced by renewed anxiety over separation. Furthermore, the toddler is no longer able to pursue goals with the indefatigability and imperviousness to setbacks that seemed almost to define the practicing period; instead, frustration and distress have taken their place. There is an inherently contradictory quality to the child's behavior during this period; on the one hand, the toddler of the rapprochement subphase strives to become increasingly independent of the mother and yet “nevertheless insistently expects the mother to share every aspect of his life” (Mahler, 1979, p. 39). Spoken words gradually replace the more developmentally primitive forms of communication, such as vocalizations and preverbal gestural language, as the toddler strives to find more efficient ways to convey his or her basic needs, affect states, and more generally, her experience of the world. At some point midway in the rapprochement subphase, the toddler's loss of the ideal sense of self in tandem with the shock of discovering that the world is not really his or her oyster ushers in the rapprochement crisis (Mahler, 1979, p. 39). As Mahler writes,

from around eighteen months on, we observed our toddlers were quite eager to exercise their growing autonomy. Increasingly, they chose not to be reminded of the times when they could not manage on their own. On the other hand, the desire to be separate, grand, and omnipotent often conflicted with the desire to have mother magically fulfill all one's desires—without the need to recognize that help was actually coming from the outside [emphasis in the original]. Thus, in a majority of cases, the prevalent mood swung to that of general dissatisfaction and insatiability, and there developed a proneness to rapid swings of mood and to temper tantrums. The period of rapprochement was thus characterized by a sometimes rapid alternation of the desire to reject mother, on the one hand, and to cling to her with coercive, determined tenacity on the other, a behavioral sequence that the word ambitendency describes most accurately… a simultaneous desire in both directions. (p. 161)

Mahler and other adherents of separation-individuation theory believed that successful resolution of the rapprochement crisis was a critical prerequisite for the development of healthy object relations and, moreover, that developmental derailments occurring during separation-individuation could lead to severe psychopathology later in life. ⁵ Although separation-individuation theory assigns a special role to the mother in promoting healthy individuation in toddlers, the influence of the father is not considered inconsequential. The father is more often on the periphery of Mahler's framework, but he is nevertheless charged with the responsibility of helping to extricate toddlers from their symbiosis with the mother. This is a task that assumes different meanings at various stages of separation-individuation, but it is probably at no time more important than during rapprochement. In James Herzog's estimation, the father not only pulls the toddler out of the maternal symbiosis but also actively intrudes on it, breaking up the “intimate, homeostatically-attuned, resonating empathy” (Herzog, 2001). During rapprochement, children develop a beginning capacity for reality testing and become more aware of gender differences, and the father's role as a “member of the family who is not mother but stands in a special relationship to the child” takes on additional meaning (Greenberg & Mitchell, 1983, p. 278).

The Subphase of Object Constancy (24–36 Months and Beyond)

Although Mahler includes this as the fourth and final subphase of her separation-individuation process, this assignation is probably somewhat misleading. In point of fact, object constancy in the course of the normal separation-individuation is first established during the third year of life, although it is most usefully thought of as an open-ended process that continually evolves throughout an individual's lifetime. ⁶ Nevertheless, it is during this early period that consolidation of object constancy occurs. Mahler thought of object constancy as consisting of two vital and interrelated tasks: the formation of stable images or mental representations of the self and of the other. In effect, she gauged the achievement of true intrapsychic separateness through the child's attainment of a more or less enduring sense of her or his own unique self, in tandem with the capacity to preserve an essentially positive, internally available image of the mother in her absence or during other times of distress (e.g., when the child experiences frustration and anger toward her). Such an achievement, in turn, required that good and bad images of both self and other become unified into self and other representations that possess both good and bad qualities.

As mentioned previously, Mahler's separation-individuation theory has been criticized on several grounds. Serious questions have been raised, for example, about the empirical soundness of her designation of the neonatal experience as “autistic,” inasmuch as newborns are responsive to bright colors, movement, and, even more significantly, to the human facial gestalt and the human voice. Furthermore, evidence has been presented in support of the existence of capacities for both short-term and long-term memory during the infant's first five months (Meltzoff & Moore, 1989; Rovee-Collier, 1987), as well as for the efforts babies undertake to establish relations between their own actions and events occurring in their physical environments (Bahrick & Watson, 1985). Collectively, such data cast serious doubt on the validity of Mahler's ideas regarding both “normal autism” and “normal symbiosis” (Fonagy & Target, 2003; Lichtenberg, 1987), assumptions that are fundamental to her view of the infant's psychological birth. Nevertheless, Mahler's ideas about separation and individuation have endured, perhaps in part because of their comparatively greater explanatory power—when compared with other psychoanalytic developmental models—in establishing a developmental basis for severe character pathology.

We are far more concerned, however, with the relationship between Mahler's ideas and those of Bowlby and with whether and in what ways they are compatible or incompatible with each other. Interestingly, Mahler's conception of rapprochement and of the rapprochement crisis, with its emphasis on the toddler's simultaneous need for maternal involvement and fear of regressive symbiotic merger, seems to parallel attachment researcher Alan Sroufe's research findings culled from his studies of “the Strange Situation” (Karen, 1998). Sroufe found that the mothers of securely attached children demonstrated the same sort of sensitivity to their toddlers as they had earlier in their children's development, adapting to the requirements of this new developmental epoch by remaining minimally though consistently engaged with the toddlers and offering just enough assistance and information without interfering with the toddlers’ quest for autonomous activity. Mahler, of course, had strongly intimated that successful passage through separation and individuation and the attainment of object constancy depended a good deal on the ability of mothers to titrate their involvement, neither withdrawing from the toddlers’ growing and at times wholly unreasonable demands nor, on the other hand, infantilizing them in response to their autonomous actions.

Bowlby, however, expressed significant reservations about a basic assumption of Mahler's, which he elaborated at some length in volume 3 of Attachment and Loss. At issue was the relationship of the child's mental representations to her or his capacity to sustain separations with emotional equanimity. In Mahler's view, libidinal object constancy was more or less coextensive with the child's capacity to endure brief separations from the mother without experiencing undue distress. In other words, the child's success in tolerating brief separations signaled the attainment of a capacity for internally accessing a positively valenced image of the mother. Bowlby, however, took exception to Mahler's position on two grounds. First, he maintained that the capacity for evocative memory typically antedates the child's ability to sustain brief separations by a year or two, sometimes longer. Furthermore, he emphasized that the representational model that a child develops of the mother is “dependent not only on the maturation of certain cognitive skills but also on the form the child's model of his mother takes, which is in turn dependent in high degree on how she treats him [emphasis added]” (Bowlby, 1980, p. 433). Such observations suggest that Bowlby was far more explicit than Mahler in acknowledging the role of the environment and of ongoing environmental transactional processes in shaping such internal mental representations. ⁷

In other, more general respects, however, Bowlby's and Mahler's views might almost be superimposed on one another. For example, Bowlby believed that there was a fundamentally complementary relationship between the infant's attachment behaviors and exploratory activity; that is, once a child had achieved attachment, attachment-seeking behaviors were extinguished and the child could then engage in exploration of the environment. Mahler commented on the very same phenomenon, although using different language, when she observed that an infant in the practicing subphase tended to seek “emotional refueling” from the mother to allay separation fears, after which exploration of the extramural landscape could begin anew (Karen, 1998). Indeed, both theorists stressed the developmental importance of parental encouragement, interest, and support at such times. Bowlby also acknowledged that although for the purposes of theory building, the differences between his formulations and those of Mahler were “of considerable consequence,” the therapeutic principles derived from Mahler's developmental scheme were “extremely close” to those Bowlby himself had derived from his theory of attachment (Bowlby, 1980, p. 433).

ERIKSON’S EPIGENETIC MODEL

A member of the small inner circle of adherents that formed around Sigmund Freud and his daughter, Anna, in the late 1920s, Erik Erikson (1902–1994) was perhaps most unlike his psychoanalytic peers. He was a talented artist who possessed an equally deep commitment to intellectual pursuits, and early on he developed a keen interest in the study of culture and its influence on the formation of personality, an interest that was further shaped through his exposure to the work of American cultural anthropologists Ruth Benedict, Margaret Mead, and Gregory Bateson. Erikson had also trained as a Montessori teacher, and he was passionate in his love of children, whether he assumed the role of educator, researcher, or psychoanalyst. ⁸

Although his contributions to psychoanalysis as a post-Freudian author are both numerous and influential, Erikson was perhaps best known for his developmental theory of psychosocial and psychosexual epigenesis. Epigenesis is a term that Erikson expropriated from the field of embryology, where it is defined as “the predetermined sequential development of the parts of an organism” (Holzman, 1998, p. 160). In fetal development, the successive emergence and predominance of each part or organ system culminates in the complex final integration of the infant's physical processes (Mitchell & Black, 1995). Erikson theorized that the ego develops in a homologous fashion, with various capacities and attributes emerging according to a sort of internal organismic timetable, and their subsequent integration into the functioning whole dependent upon the successful completion of the stage of development that has immediately preceded it. Erikson also believed that in such a developmental schema, earlier stages of growth are inherently more vulnerable to environmental disruption than are later ones; furthermore, more profound developmental sequelae would be likely to follow such early disruptions, affecting all subsequent later developmental stages.

Erikson's psychosocial/psychosexual epigenetic theory charted ego development across the entire life span and highlighted social-environmental factors to a greater degree than had any previous psychoanalytic developmental model. His most enduring work, Childhood and Society (1950), is perhaps the best illustration of his profound interest in the way individual development is shaped by the influences of culture, or the “interpenetrability of individual and culture” (Mitchell & Black, 1995, p. 143). In this work, the final shape of the individual personality is deemed to be subject to a variety of influences aside from the purely libidinal interests that Freud had identified; such influences were culture-specific and could also involve complex geographical, economic, and social forces only hinted at by earlier psychoanalytic authors. ⁹

In Erikson's view, healthy ego development depends upon the mastery of specific developmental tasks and normative crises associated with each of the eight life cycle stages that he identified (see Figure 1). Citing the tendency to misinterpret and oversimplify Erikson's complex developmental timetable, some have characterized the “crisis” of each successive developmental epoch as more accurately representing a sort of dialectical tension—reinforcing the notion that trust and mistrust, autonomy and shame and doubt, industry and inferiority, and so forth are always in a complementary relation to each other.

Figure 1 Erikson's eight-stage epigenetic model. (Source: Adapted from Perlman & Brandell, 2010, pp. 56–57.)

Even though one or another crisis is in the forefront at any particular time, all these issues and tensions are active throughout the life cycle. Each stage is reworked anew by the struggle with subsequent ego qualities, and Erikson envisioned ego development across life cycle stages less in terms of a stepladder and more in terms of a complex set of tensions, progressively unfolding and in constant resonance with each other. (Mitchell & Black, 1995, p. 148)

Although many writers have emphasized Erikson's unique contributions to our understanding of adolescence and to the stages of adult development—largely uncharted terrain when the first edition of Childhood and Society appeared—his conception of early childhood is equally radical. While preserving the essential structure of the classical theory of libidinal stages, Erikson's vision of the oral, anal, and phallic stages was transformed by a concomitant emphasis on both the historical and the contemporary influences of culture. In the Eriksonian framework, in marked contrast not only to classical conceptions but also to then-prevailing views of ego psychology, cultural forces were thrust into the foreground, possessing causality and capable of generating meanings of their own (Mitchell & Black, 1995). Although Erikson never disputed the common ground that the erogenous zone/libidinal stage theory offered, he was equally determined to highlight the dialectical relationship that existed between the individual and the culture of which he or she was a part, which he once termed “a psychosocial relativity” (Erikson, 1968, p. 23). Obviously, developmental deviations and psychopathology could also then not be presented in absolute terms. Consider the following passages from Childhood and Society:

It is already in his earliest encounters that the human infant meets up with the basic modalities of his culture. … The pathology and irrationality of oral trends depend entirely on the degree to which they are integrated with the rest of the personality and the degree to which they fit the general cultural pattern. … Here, as elsewhere, we must therefore consider as a topic for discussion the expression of infantile urges in cultural patterns which one may (or may not) consider a pathological deviation in the total economic or moral system of a culture or nation [emphasis added].(Erikson, 1950, pp. 60–64)

Erikson also focused to a much greater degree than had Freud on the nature of the interpersonal processes that characterized the earliest relations of infants and mothers. Furthermore, his concept of “organ modes” or “modes of functioning” furnished a basis for meaningfully extending psychical aspects of bodily functions. Hungry infants, Erikson wrote, are not only interested in being nursed or fed; they are receptive in many other ways, as well; they “take in” with their eyes, as well as via tactile sensations, anything that feels good. Toddlers, in contradistinction, struggle with the dialectic of “holding on” and “letting go,” oedipal-age children with “intrusiveness” and “inclusiveness,” and so on (Erikson, 1950, 1997).

Although Erikson devoted somewhat less attention to separation and attachment, one might argue that his ideas in this area are implied in his focus on breast-feeding and weaning and, more generally, in his depiction of the maternal-child relationship during the oral-respiratory-kinesthetic stage, with its psychosocial crisis of basic trust versus basic mistrust. This crisis, which Erikson dates to the second half of the first year, is ushered in by three more or less concurrent developments. The first, he observes, is physiological, consisting of the “general tension associated with a more violent drive to incorporate, appropriate, and observe more actively” and is linked to teething; the second is psychological, involving a gradually accruing awareness on the infant's part of herself or himself as a separate person. The third, however, is “environmental,” consisting of the mother's “apparent turning away from the baby” in order to resume various work and/or social activities suspended since the beginning of her pregnancy (Erikson, 1959, p. 60). Although all of these developments are considered normative, Erikson suggests that the sudden physical loss of the infant's mother “without proper substitution” carries with it the risk of anaclitic (infantile) depression as well as the potential for lasting psychological damage.

Bowlby believed, however, that Erikson, despite his tremendous creativity in integrating theories of drives with post-Freudian conceptions of ego and sociocultural forces, seemed intent on preserving both Freud's original erogenous zone framework and with it an apparently irreducible relationship between orality and attachment. The relatedness of infant and mother, which Erikson freely acknowledges to be of such importance in the earliest months of life, is not subject to a secondary drive, but rather is fundamentally linked to orality: “The oral stages then, form in the infant the springs of the basic sense of trust” (Erikson, 1950, p. 75). In Bowlby's estimation (Bowlby, 1969), the importance that Erikson attributes to “primary object sucking” in this first developmental period, in tandem with his failure, generally speaking, to account more fully for attachment phenomena within the psychosocial epigenetic theory, places him in a position that is just as untenable as those of Sullivan and René Spitz (another infant developmentalist). All three, Bowlby notes, are “trapped” by the disparity between “their clinical appreciation of the facts” and their “conventional theorizing” (Bowlby, 1969, p. 374).

Bowlby, however, may not have been completely fair in his dismissive critique of Erikson on the grounds just cited. Indeed, just nine years after the publication of Childhood and Society, Erikson appeared to have placed somewhat more emphasis on the qualitative aspects of the infant-mother relationship, declaring, “[The] amount of trust derived from earliest infantile experience does not seem to depend on absolute quantities of food or demonstrations of love, but rather on the quality of the maternal relationship” (1959, p. 63). Erikson's important and pioneering ideas regarding “reciprocity or mutual regulation,” first described in Childhood and Society (1950, p. 58), seem to parallel the notion of interactional synchrony, a concept attributed to attachment theorists (Isabella & Belsky, 1991). Furthermore, his interest in a life span perspective also represents an important source of convergence with Bowlby's framework. Then, too, there was the recognition Erikson gave to cultural forces, and to the immediate social context of mothering, both of which are resonant with much current thinking in the attachment field (Fonagy, 2001). Finally, Erikson's ideas regarding the longer-term sequelae of trust and mistrust, and specifically trust and identity, are upon closer examination rather close to Main's concept of coherence, a central organizing theme in her framework for adult attachment (Fonagy, 2001). Such similarities have led some to conclude that Erikson's ideas, “in both spirit and particulars,” demonstrate a consistency “with contemporary research on infant attachment and its roots in interactions with the primary caregiver” (Fonagy, 2001, p. 59).

KOHUT’S PSYCHOLOGY OF THE SELF

A relative newcomer among the psychoanalytic psychologies that comprise contemporary psychoanalysis (classical psychoanalysis, ego psychology, and object relations theories, among others), the psychology of the self ¹⁰ was introduced by the American psychoanalyst Heinz Kohut (1913–1981) in a series of essays and books published between 1959 and 1984. Although originally presented within the existing classical drive theory (Kohut, 1966, 1971), Kohut's ideas about development, psychopathology, and psychoanalytic treatment were gradually expanded and revised, culminating in a distinctive and fundamentally new psychoanalytic psychology (Kohut, 1977, 1984). The essence of self-psychology may well lie in its vision of the human condition. Freud's view of humankind is usually expressed in terms of conflicted desire—an ongoing battle between libidinal or aggressive wishes on the one hand and societal precepts on the other. Within this framework, the individual's capacity to experience guilt, however painful it may prove to be, offers the only certain means for the renunciation of instinct, without which capacity a civilized society is not possible.

Kohut's perspective on the human condition, however, appears to contrast markedly with that of Freud. Kohut was far less interested in primitive desires and the conflicts they engendered than he was in the loss of meaning associated with contemporary life—i.e., in those experiences that might lead to a “fractured, enfeebled, discontinuous human existence” (Kohut, 1977, p. 238). In classical theory, man lives within the “pleasure principle,” endeavoring “to satisfy his pleasure-seeking drives,” whereas self-psychology emphasizes human failures, uncompleted projects, and unrequited efforts at work and love (p. 238).

Kohut and his followers, much like the principal figures associated with other psychoanalytic psychologies, developed not only a framework for clinical interventions but also a model for understanding various aspects of normative human development/developmental deviations, and a theory of psychopathology. However, self-psychology has also developed its own lexicon, one sufficiently different from that of other psychoanalytic psychologies to warrant a brief review of several of its key concepts and terms:

Selfobjects

In self-psychology, the term selfobject is used to represent a particular kind of object relationship in which the object is experienced as being a part of or extension of the subject's self, and in which little or no psychological differentiation occurs. Kohut observed that the control one expects over such selfobjects is roughly equivalent to the sort of control an adult “expects to have over his own body and mind,” in contradistinction to the “control which he expects to have over others” (Kohut, 1971, pp. 26–27). He believed that human beings require three distinctly different kinds of selfobject experiences: (1) mirroring, (2) idealizing, and (3) partnering, each of which, assuming optimal conditions for development, is furnished to the child in an attuned, empathically resonant interpersonal milieu. Mirroring selfobjects “respond to and confirm the child's innate sense of vigor, greatness, and perfection,” whereas idealized selfobjects offer the child the powerful and reassuring presence of caregivers “to whom the child can look up and with whom he can merge as an image of calmness, infallibility, and omnipotence” (Kohut & Wolf, 1978, p. 414). Finally, partnering objects provide the child with various opportunities through which a sense of belonging and of essential alikeness within a community of others may be acquired.

The Tripolar Self

Kohut's theory identified an intrapsychic structure, the tripolar self, that links each selfobject relationship to a corresponding sphere of self experience. Mirroring experiences, which reflect the need for approbation, interest, and affirmation, are associated with the grandiose exhibitionistic self; idealizing experiences, which reflect the need for closeness and support from an (omnipotent) idealized other, are linked to the intrapsychic structure known as the idealized parent imago; and partnering, associated with the individual's need for contact with others who are experienced as having an essential similarity to oneself, is linked to the alter ego. In the course of development, these three poles crystallize as the result of various needs of the evolving self, as well as the responses of significant people in the child's environment who serve as selfobjects (Leider, 1996).

Empathy and Transmuting Internalization

According to the self-psychologists, empathic processes originate in early infancy and traverse the entire life span, and breaches or traumatic disruptions in the empathic attunement between self and selfobject are believed to possess special importance. Self-psychology defines empathy as “vicarious introspection” or the feeling of oneself into the experience of an other. The capacity of parents or parental surrogates in the child's selfobject milieu to furnish empathically attuned responses is considered a critical sine qua non for the healthy development of the self. However, somewhat paradoxically, self-psychological theory also maintains that comparatively minor, nontraumatic lapses in parental empathy are equally important as catalysts for the development of what Kohut termed transmuting internalizations—an incremental “taking in” or internalizing of various functions originally associated with the selfobjects. These functions and attributes, which include self-soothing and self-calming, pride, wisdom, humor, stoicism in the face of adversity, and indefatigability in the pursuit of personal goals, are absorbed and metabolized through a virtually imperceptible, bit-by-bit process of translocation. Ultimately, they become enduring components of the child's own self-structure, altered by his or her individual imprimatur.

Essential elements contributing to transmuting internalizations are sequenced in the following order:

1. optimal frustration,

2. increased tension,

3. selfobject response,

4. reduced tension,

5. memory trace,

6. development of internal regulating (self) structures.

The Self Types

Self-psychologists believe that the self is most usefully understood within the intersecting matrices of developmental level and structural state; they have identified four principal self types:

1. the virtual self, an image of the newborn's self that originally exists within the parent's mind, evolving in particular ways as the parental “selfobjects empathically respond to certain potentialities of the child” (Kohut, 1977, p. 100);

2. the nuclear self, a core self that emerges in the infant's second year and serves as the basis for the child's “sense of being an independent center of initiative and perception” (Kohut, 1977, p. 177);

3. the cohesive self, the basic self structure of a healthily functioning, well-adapted individual, reflecting the harmonious “interplay of ambitions, ideals, and talents with the opportunities of everyday reality” (Leider, 1996, p. 143); and

4. the grandiose self, a normal self structure of infancy and early childhood that comes into existence in response to the selfobject's attunement with the child's sense of himself or herself as the center of the universe.

Kohut's ideas about development and psychopathology have been criticized on a number of different grounds. His concept of grandiosity as representing a normal stage of infantile development and his notion of infantile omnipotence have both been challenged, and limited evidence exists to support the child's need for unconditional admiration; furthermore, there is little research evidence confirming the existence of a narcissistic period of development in the child's first two to three years (Fonagy & Target, 2003). However, other self-psychological ideas, such as the mother's role in imparting tension-regulation skills to her baby and the relationship of transmuting internalizations to the building up of internal self structures, have received a greater degree of support (Fonagy & Target, 2003).¹⁰

Although Kohut's discussion of early infantile development focuses on the origin and evolution of the infant's self structure, his ideas regarding development do not appear to be fundamentally incompatible with attachment theory and in fact may possess a certain complementarity with it. While he does not offer a developmental timetable equivalent to others reviewed here, he strongly intimates that specific kinds of selfobject responses prove either facilitative or problematic in the child's quest for the acquisition of stable and enduring self structures and that such experiences are associated with particular developmental epochs. For example, grandiosity is transformed into exhibitionism via maternal mirroring between the second and fourth years, and idealized goals make their appearance sometime between the fourth and sixth years (Fonagy & Target, 2003). Of somewhat greater interest, however, is Kohut's vision of the baby's earliest development, specifically his conception of both merger experiences and idealizing experiences.

The basic configuration for Kohut's model of idealizing experiences, i.e., self/selfobject encounters with an empathically resonant, calming, reassuring (and from the infant's point of view, omnipotent) selfobject, is that of “uplifting care.” Although Kohut employs this phrase in both a literal and a figurative sense, he frequently takes note of the physical dimension of attuned caregiving. At the same time, however, the caregiver's provision of such “uplifting care” is but one aspect of the empathic-responsive human milieu into which each child is born. Kohut acknowledges the importance of the child's physical attachment to her mother when “she is picked up by mother and thereby feels herself part of the omnipotent strength and calmness of the idealized selfobject” (Kohut, 1984, p. 186), yet he suggests that such experiences are most usefully understood within the framework of the child's efforts to acquire intrapsychic self structure. Physical attachment, in Kohut's system, represents but one variety of selfobject encounter—granted, a prototypical one; his real focus is on the role of empathic processes or the lack thereof in shaping such idealizing experiences. Put somewhat differently, secure attachment, from the perspective of self-psychology, is a result of optimally functioning maternal empathy rather than a subject of study unto itself. Conversely, the self-psychologist's interest in insecure patterns of attachment (anxious, disorganized, and so forth) would likely center on the traumatic breaches that gave rise to such patterns and which specific realm of selfobject experience was involved. These assumptions may also reveal a more fundamental difference between self-psychology and attachment theory: for Kohut, the principal motivating force behind human behavior is that of self-cohesion, and the most basic anxiety that of disintegration—the individual's experience of “a defect, a lack of cohesiveness and continuity in the sense of self” (Fonagy, 2001, p. 110). By contrast, attachment theory focuses on a biologically predefined pattern of relationship, one in which the attachment figure remains very much at the center of the equation (Fonagy, 2001, p. 110).

Nevertheless, important points of basic convergence between attachment theory and psychoanalytic self-psychology remain. Both Bowlby and Kohut believed that Freud's erogenous zone/libidinal phase model provided an inadequate account of human development, and both attributed importance to the baby's physical attachment to the mother, as well as to the overriding importance of maternal-infant attunement. Moreover, both placed considerable emphasis on the role of maternal regulating functions and on the potentially traumatogenic effects of disruptions in the physical and, later, emotional relationship between infant and caregiver.

CONCLUSION

This paper has examined four different contemporary psychoanalytic models of development and their “goodness of fit” with attachment theory. Although important differences exist between Bowlby's ideas and those of Sullivan, Mahler, Erikson, and Kohut, there were, surprisingly, points of theoretical convergence in virtually every case. Margaret Mahler's separation-individuation theory, for example, emphasizes the complementary relationship of exploratory behavior to attachment security, a theme common also to Bowlby's theory. Sullivan's interpersonal conception of development, like Bowlby's ideas about attachment, places considerable emphasis on environmental factors both in shaping the personality and in psychopathology. Kohut's focus on the significance of early merger states and, more generally, on maternal attunement in the formation of healthy self structure also appears compatible with Bowlby's views. And finally, Erikson's epigenetic theory and his conception of basic trust in particular, despite Bowlby's critique of his early adherence to the basic outline of the classical erotogenic zone model and his equation of orality with attachment, demonstrates a meaningful convergence with the Bowlbyan framework.

Notes

1. The close parallel between Sullivan's use of this term and the social work paradigm of person-in-environment seems unmistakable here, although Sullivan has not often been cited by social work historians as having exerted significant influence on the evolution of social work theory.

2. This “need,” however, is put forward far more tentatively than many of Sullivan's other ideas about early development, a fact that did not escape the notice of John Bowlby (1969, pp. 374–375).

3. Researchers in the field of infant development have challenged such ideas, as the upcoming discussion of Margaret Mahler's separation-individuation theory will highlight.

4. This was particularly true with respect to the timing and characteristics of the phases (Greenberg & Mitchell, 1983).

5. Masterson and Rinsley (Masterson, 1976, 1985; Masterson & Rinsley, 1975; Rinsley, 1977, 1978) are perhaps best known for their distillation of separation-individuation theory and their application of it to the understanding and treatment of borderline, schizoid, and other character-disordered clients.

6. Such libidinal object constancy is, however, not the equivalent of the Piagetian object permanence; the latter represents a milestone in cognitive development held to occur at 18 to 20 months and concerned primarily with the child's relationships to the world of inanimate objects.

7. Bowlby had earlier (1973) taken exception to Mahler's (1968) assertion that separation anxiety was possible only after completion of separation-individuation and the attainment of object constancy. However, Mahler's use of this term gradually expanded to include infants in not only the rapprochement but also the practicing subphase (1978), which would appear to address Bowlby's objections.

8. Dr. Martin Loeb (deceased), a cultural anthropologist and social worker, shared an office with Erikson in the 1950s. He once recalled that one of Erikson's greatest pleasures was simply to watch the children at play in the schoolyard across the street from their office building, a pursuit in which he spent countless hours (personal communication to J. Brandell).

9. Erikson's detailed description and comparative analysis of child-rearing practices of the North American Sioux and Yurok tribes in Childhood and Society is one such example.

10. The reader is referred to Fonagy and Target (2003) for additional discussion of the scientific status of self-psychology.