An Integrated Developmental Approach to Personality Disorders in Adolescence: Expanding Kernberg’s Object Relations Theory

Dr. Otto F. Kernberg elaborated a paradigm for understanding the development and treatment of severe personality disorders that has been prescient and has had important implications for understanding and identifying personality disorders among adolescents. He has considered identity diffusion to play the central role in personality pathologies, a view that has now been widely integrated into the field, including into the Alternative DSM-5 Model for Personality Disorders (1), which emphasizes the centrality of identity, interpersonal functioning, and investment in work or school to the concept of personality. His insights about borderline personality disorder have led to the broader construct of borderline personality organization, which differentiates severities of personality pathology as well as types of personality disorders (2). The elaboration of his object relations model of borderline personality disorders has been accompanied by the development of transference-focused psychotherapy (TFP) for borderline personality disorders (3), conceptualizations of higher-level personality pathology (4), and TFP for narcissistic personality disorders (5).

Furthermore, working within a psychoanalytic tradition, Kernberg has used research to develop his theories and has used neurobiological findings to elaborate his model of personality pathology (6). This work has included an integration of Panksepp’s neurobiological theory of affects (7, 8), reflected in Kernberg’s theorizing of how the positive affect systems of attachment, eroticism, and play-bonding give rise to the fundamental motivational drive for dependency, sexuality, and affiliative interactions, which jointly constitute the libido. Kernberg, in his work with Silbersweig and colleagues (9), has also utilized MRI technology to investigate neural bases for difficulties in behavioral inhibition associated with borderline personality disorder.

This article presents Kernberg’s model of the development of personality disorders in adolescence. He emphasizes the centrality of internalized object relations in identity formation during this age period and provides a framework for the assessment of personality organization and the applicability of TFP to this age group (10).

An Object Relations Model of Psychological Development and Identity

Drawing on Melanie Klein’s object relations theory (11), Kernberg theorizes that the child, from infancy onward, forms internal mental representations of the self in relation to others or attachment figures who are considered the object of the child’s needs and wishes. In Kernberg’s model (12, 13), affects play a central role in the development of internal object relations. Affects are the basic building blocks of the psychological structures that come to constitute identity and organize an individual’s affective life, motivations, desires (especially toward attachment figures), and habitual behaviors. Internalized object relations are theorized to be composed of affects linked to representations of a specific interaction between the self and an other. Examples include affectionate feelings linked to the representation of oneself as a loved child and to an other as a caring mother; the affect of rage linked to the self-representation of being a coerced adolescent in relation to a powerful and inflexible parental figure; and shame linked to the self-image of an enamored, vulnerable adolescent relating to an other who is indifferent and uncaring toward the adolescent.

Use of the concept of internal representations of the self and others is not unique to Kernberg; it is also central to the attachment framework. For example, Bretherton and Munholland (14), when describing John Bowlby’s concept of internal working models (IWMs) in attachment behavior, see internal representations of the self and others as a “dynamic representation system” of self and others based on early responsiveness of attachment figures that comes to be used to “imagine habitually experienced social interactions.”

Developmental Theory During Infancy and Early Childhood

In Kernberg’s theory (15–17), internal object relations are derived from the interplay between the baby’s temperament and genetic affective predispositions, in interaction with an attachment figure (usually the mother). Consistent with attachment theory and research, early mother-infant interactions are theorized to produce affective procedural memories that are assimilated into “memory traces” of positively or negatively charged representations of the self in interaction with an other. These positive and negative affective procedural memories are theorized to develop separately at first but later actively split or become dissociated from each other to maintain “ideal all-good experiences” and to avoid “persecutory all-bad experiences.” As children develop cognitively and become increasingly able to explore and assess external reality, all-good and all-bad experiences are integrated because the child increasingly becomes aware that the attachment figures have both good and bad aspects. It is theorized that if the child has a predominance of positive experiences, the negative experiences become more tolerable and are progressively integrated into a comprehensive, realistic view of the self and other. Influenced by Mahler (18), Kernberg sees an early primitive integrated self as crystalizing between ages 3 and 5 (Kernberg, 2012, unpublished manuscript). When this process fails, the child’s sense of identity becomes diffuse, which Kernberg associates with risk of developing a personality disorder. Risk factors for the development of identity diffusion include negative affect, impulsivity, and deficits of effortful control (19–21), which may occur in combination with exposure to physical or sexual trauma, parental antipathy, abandonment, or chronic family chaos characteristic of disorganized attachment (21).

Theorizing and conducting research on attachment have provided concurrent support for the concepts that are fundamental to Kernberg’s object relations model. For example, the formation of IWMs begins early in life, before speech develops. As cognitive, language, and social abilities develop, IWMs are revised and become increasingly more complex (14). As with the object relations dyad, Bretherton and Munholland (22) note that attachment relations are never affectively neutral. Furthermore, the importance given to early experiences of parenting in the object relations model is consistent with convergent findings from attachment research showing that differences in maternal sensitivity and responsiveness are associated with distinct attachment styles in children as young as 15 months (23). A vivid example is provided by Lyons-Ruth and Jacobvitz’s work with children with disorganized attachment styles (24). Infants with disorganized attachment manifest behaviors during the reunion phase of the strange situation paradigm that do not correspond to one of three more typical organized attachment patterns. These behaviors are indicative of “contradictory intentions” (e.g., approaching parents with the head averted) that present as confusing combinations of approach and avoidance. Disorganized infant attachment was first described in infants of mothers who had experienced recent bereavement. It is also frequently associated with abuse or neglect, in which case the attachment figure whom the child turns to for safety when distressed is also the source of fear, thus leaving the infant caught in an impossible bind. These descriptions are consistent with formulations of the early development of splitting. Furthermore, Lyons-Ruth and Jacobvitz’s work indicates that disorganized attachment among infants is also associated with hostile or helpless parenting styles, where the parent fails to provide the minimally expected responsiveness to infant distress. In their review, Lyons-Ruth and Jacobvitz (24) note that infants with disorganized attachment also show a variety of disturbances in affective communication and develop IWMs characterized by emotional incoherence. For Kernberg and attachment theorists, these internal mental representations undergo elaboration and modification on the basis of the subsequent emergence of new capacities and experiences with others.

Identity Consolidation, Identity Crisis, and Identity Diffusion

Identity consolidation is classically seen as a hallmark of healthy adolescent development and, in Kernberg’s model (Kernberg, 2021, personal communication to L.N.), is associated with typical reorganization of the infantile self-other representations and with the process of expanding identity toward an integrated sexuality, maturation in moral development, changes in ego ideal, and autonomy. Drawing on Erikson’s distinction between typical ego identity, identity crisis, and identity diffusion and on his notion of identity integration as underlying typical personality development (25, 26), Kernberg proposed that severe personality pathology is characterized by identity diffusion (15, 27). Furthermore, Kernberg proposed that identity diffusion underlies borderline personality disorder and the characteristic difficulties associated with this disorder: turbulent interpersonal relationships and interpersonal difficulties, emotional dysregulation, and impulsive behaviors such as suicidality and self-harm. In Kernberg’s conceptualization, identity diffusion is characterized by impoverished, split, or extreme representations of self and others so that the individual has no real sense of their unique self or of the personal qualities of themselves or others, including their strengths and weaknesses. This identity state is associated with an unstable sense of self and a lack of a sense of others as people who contribute both positively and negatively to relationships. Furthermore, identity diffusion is associated with primitive defenses, including splitting, dissociation, and projective identification. In contrast, the development of an integrated identity is characterized by a stable and realistic sense of self and others, where others are seen as individuals with their own unique qualities, motivations, and intentions. Therefore, an adolescent’s level of identity integration is an important indicator of their personality organization and distinguishes between identity diffusion (indicative of personality disorders), identity crisis, and typical development. Furthermore, identity integration can be used as an indicator to monitor progress in psychotherapy.

Developmental Theory During Late Childhood and Adolescence

Biopsychosocial processes specific to adolescence challenge the adolescent to expand representations of the self and others and to develop increased concern and empathy for and awareness of others. This expansion is supported by cognitive maturation and is driven by neurobiological and hormonal changes that result in physical and sexual maturation and intensified aggressive and sexual impulses (28). Furthermore, the broadening of the adolescent’s social life, friendships, and love relationships challenges them to reexamine parental value systems, clarify their own values, integrate new idealized role models, and clarify their own moral system and ego ideal. This process of maturation is considered to continue until young adulthood, when a coherent and mature identity is usually attained in the context of typical development. These transformations, for many, are associated with a “normal” trajectory marked by effective adaptations to work, relationships, and a sense of self, whereas, for others, these transformations may result in identity crisis or identity diffusion and the emergence and intensification of a personality disorder. We describe each of these adolescent pathways below.

Typical Identity Development in Adolescence

The development of an integrated identity in adolescence is characterized by four central features. First, the infantile identity is reorganized into a more mature identity, which integrates sexuality and aggressive impulses activated by puberty-associated hormonal changes, along with changes in the internal world of object relations that come with an expansion of social life and integration of new adult role models and intimate relationships. Second, more mature defenses, such as repression, rationalization, and humor, are used. Third, the superego and direction of sexual impulses are reorganized and attachment relationships outside the family are formed. Fourth, the ego ideal is modified and enriched through the integration of new idealized figures and aspirations (29–32).

Identity Crisis in Adolescence

Kernberg (10, 27) defines an identity crisis in the adolescent as a period that features a lack of correspondence or a disjunction between how the adolescent sees themself and how they feel seen and treated by others. An identity crisis must be differentiated from typical adolescent identity development as well as from identity diffusion associated with personality disorders (33). The adolescent’s personal history within the family as well as the family’s and others’ inability to acknowledge the changes within the adolescent contribute to the crisis. Therefore, identity crises derive from a lack of confirmation by others of the adolescent’s changing identity, but the adolescent’s capacity for in-depth object relations remains intact. This capacity distinguishes adolescent crisis from identity diffusion. For example, despite being rebellious and having conflicts with parents, the adolescent maintains a sense of self and the capacity to see the positive and negative aspects of their parents. This capacity is illustrated by a comment to his therapist of a 16-year-old adolescent who repeatedly clashed with his mother, who had borderline personality disorder, and his father, who would often capitulate to her: “I’m pushing out for more air than I want to, so I’ll have enough air.” This comment was elicited in anticipation of his mother’s pushback and his parents’ unwillingness to recognize and accept any differences between him and themselves or his siblings.

Identity Diffusion in Adolescence

Several features distinguish identity diffusion from typical adolescent development: the dominance of primitive defenses such as splitting, projective identification, and omnipotent control; a poorly integrated value system; a lack of a clear sense of self or contradictory self-experiences; a poor sense of others and incapacity to have a stable, realistic, and reasonably balanced sense of others, as revealed in multiple conflicts at school, at home, and at work; a lack of realistic goals, motivation, and investment in learning that often contributes to school failure; chaotic sexual behaviors; and an incapacity to function autonomously while maintaining healthy dependency on attachment figures (34, 35).

It is remarkable testament to Kernberg that identity has now gained a central position in the diagnosis of personality disorders in general and that the features elaborated by Kernberg over the years have largely become incorporated into the Alternative DSM-5 Model for Personality Disorders (1). For example, having a poor sense of others and incapacity to have a stable sense of others, described above, can be seen in the Alternative DSM-5 Model for Personality Disorders description of identity. Having a lack of realistic goals, motivation, and investment in learning can be seen in the DSM-5’s interpersonal element of empathy and in the domain of self-direction.

When an individual enters the adolescent period with an already well-integrated identity, remodeling of identity usually proceeds smoothly through the reorganization of self-identity, the assimilation of novel experiences with new significant others, and the integration of sexuality and love. This process is further facilitated by parents who support the adolescent’s normative push for autonomy and independence, tolerate their momentary regressive phases, and maintain realistic expectations and flexible rules. However, preexisting identity diffusion increases the risk of regression.

In summary, Kernberg proposes that an expansion of identity occurs during adolescence and is necessary for identity integration. Although there is a normative push for identity expansion during adolescence, a fear of expansion and autonomy may result in premature consolidation consistent with the moratorium in identity development described by Marcia (36) but formulated by Kernberg within an object relations model.

Assessment of Personality Disorders in Adolescence

Another important contribution to the understanding of personality disorders by Kernberg and his group has been the development of two assessment tools, which were initially developed for adults (37) and have since been adapted for use with adolescents. The first, the Structural Interview (38), is an open-ended clinical interview used to diagnose borderline personality organization. The second, the Structured Interview of Personality Organization in Adolescence (STIPO-A; unpublished document, Personality Disorders Institute), was developed to assess the depth and level of identity integration within adolescent personality organization (39). The reliability and validity of the latest version of the STIPO-A is currently being examined.

Structural Interview

The format of the Structural Interview is more open ended than structured or semistructured interviewing, and the interview is guided in part by the adolescent’s responses and manner of self-presentation. The interviewer might begin with a question designed to explore the adolescent’s understanding of their current problems, such as, “Tell me, what brought you here? What is your understanding of why we are meeting?” The interview then progresses toward the exploration of the adolescent’s awareness and understanding of their problems, their wish for treatment, and their motivation for change. Responses to the prompt, “Tell me about yourself” provide more insight into the young person’s identity status. This inquiry can be expanded by asking them to describe others (e.g., parents, best friend) (40) and, if necessary, can be made more concrete (e.g., “How might you describe yourself to someone else your age so that they would have a good picture of who you are?”).

Three adolescents’ responses to the prompt, “Tell me about yourself” are presented below to illustrate different levels of borderline personality organization on the basis of each adolescent’s level of integration of self as well as their defensive style.

Case 1.

Jenny (identifying information has been changed for all case descriptions), an intelligent, dramatic 16-year-old girl who was hospitalized for self-mutilation, had a friendly demeanor and quickly began to talk about her mother without being prompted to do so. However, when the interviewer said, “Now, tell me about yourself,” she recoiled in evident panic, had difficulty remaining seated, and was unable to talk about herself. “I am used to questions; please ask me questions,” she said. Jenny’s inability to describe herself reflects the absence of mentalizing regarding herself and is indicative of a disorganized inner world, reflecting identity diffusion and rendering her dependent on others to provide her with an organizing external structure.

Case 2.

John, a 13-year-old boy, was referred because he attacked his peers and teacher. He reported having undergone an involuntary psychiatric hold for doing “rude” things to a boy, but then suddenly displayed a defensive role reversal, claiming that he was the victim and was attacked by the other boy. When prompted, “Tell me about yourself,” he gave vague and confusing replies. He attempted to reorganize himself by focusing on the other boy rather than on himself and by reversing roles, commenting that the boy was “delusional.” When asked whether he had any delusions, he quickly expressed the fear that others would attack him. After a pause, the interviewer said, “Tell me if I followed you correctly. You thought the other boy believed you were a threat to him and so he attacked you, and you often believe people want to harm you and so you attack them.” He looked intently at the therapist and nodded. When the therapist added, “So, is it possible that you attack people because you believe they want to destroy you and it makes you feel safe if you destroy them first?” he sighed and said, “Finally, someone understands.”

Case 3.

Claire, a 15-year-old girl, had made two suicide attempts and had multiple cuts on both arms and significant weight loss. She did not respond well to pharmacological approaches to treating her depression and seemed disengaged and barely communicative with her therapist and hospital staff. She reported that she hurts herself but gave vague and cursory replies with a flat tone and no eye contact. When asked to spontaneously talk about herself, she described her current feeling of depression and added that she used to be happy, outgoing, and even loud 4 years ago. This surprising revelation and the striking contrast between her current self and her previous self, which she liked but was disconnected from, seemed to be more characteristic of severe splitting than depressive disorder. She revealed that she suppresses this positive part of herself that she likes and misses and that she often feels intense anger toward several people in her life. She said that she could not show that anger or hit those people, so her only option was to cut herself. It became clear that her depression masked her rage. Although she maintained her flat tone and somewhat distant overt demeanor, she became more engaged with the therapist and responsive when the focus of the discussion shifted from her symptoms to herself. Claire showed a more nuanced sense of herself than did Jenny or John, suggesting that Claire was at a higher level of borderline personality organization.

STIPO-A

The STIPO-A (unpublished document, Personality Disorders Institute) based on the Structured Interview of Personality Organization–Revised for adults (39), was developed to evaluate domains central to Kernberg’s notion of identity integration as a feature of typical adolescent personality development. The STIPO-A operationalizes personality organization for assessment and research purposes. The domains covered by the STIPO-A include identity (e.g., depth of representation of self, consistency of self-representation across relationships, autonomy, and dependency), object relations (e.g., portrayal of family and friendships), reality testing, defenses (e.g., repression [higher level] or splitting [lower level]), aggression (e.g., self-directed [neglect, self-harm, suicidality] or other directed [temper, attack, sadism]), and moral functioning (e.g., internal values, rule breaking, exploitation). The assessment includes questions about puberty and sexuality (e.g., gender identity, body image, experience of sexuality, intimacy) as well as narcissism (e.g., grandiosity and vulnerability, ego ideal, ideal self) (41). Adolescents respond well to the interview, which provides a profile of personality organization that informs diagnosis and treatment choice, including the appropriateness of TFP for adolescents (10).

Conclusions

Otto F. Kernberg’s legacy is evident in the current understanding of personality disorders in terms of identity diffusion and the view of identity integration as being associated with balanced representations of self and others. Pioneers like Paulina F. Kernberg (40, 42, 43) first drew attention to the early manifestations of personality disorders in children and adolescents, and Chanen et al. (44–46) and Sharp et al. (47–50) made major contributions to recognizing personality disorders in adolescence. Kernberg’s personality organization framework is especially pertinent for assessing personality disorders and for differentiating typical development from personality pathology among adolescents.