Otto F. Kernberg

Kernberg was born Otto Friedmann Kernberg in Vienna on 10 September 1928 to Jewish parents, Leon and Sonia Paula Kernberg. After Nazi Germany annexed Austria he was expelled from his school, and in 1939 the family fled to Chile. He studied biology and medicine at the University of Chile. He trained in psychiatry and psychoanalysis at the Chilean Psychoanalytic Society, which Ignacio Matte Blanco had founded. In an interview he described the training as an integration of classical descriptive German psychiatry with psychoanalytic and psychodynamic psychiatry.

He recalled a patient he treated as a candidate in a supervised analysis in Chile. The man had been diagnosed as an obsessive-compulsive personality and did not change over years of treatment, and the memory stayed with him. Later, Hermann van der Waals, who had written on the narcissistic personality, told him that patients of this kind were narcissistic personalities. Kernberg dated his own work on narcissistic pathology from that recognition, remarking that nobody had yet described these characteristics well in the literature.

In 1959 he went to the United States on a Rockefeller Foundation fellowship to study psychotherapy research with Jerome Frank at Johns Hopkins Hospital. There he met John White, whose method of clinical interviewing, he said, inspired his later structural interviewing.

He was married to the child psychiatrist Paulina Kernberg until her death in 2006. In 2008 he married the psychologist Catherine Haran.

At the Menninger Foundation in Topeka, Kansas, Kernberg joined, and eventually directed, the Psychotherapy Research Project. He was also a supervising and training analyst at the Topeka Institute for Psychoanalysis. By his own account the project, begun in 1954, followed 42 patients. Half were treated in standard psychoanalysis and half in psychotherapies with a psychoanalytic basis, and about half of them had severe borderline conditions. Every session had typed process notes, kept over years of treatment, and he called the material a gold mine.

From this clinical and research work came a series of papers in the late 1960s and early 1970s. They include Borderline Personality Organization¹ (1967) and Factors in the Psychoanalytic Treatment of Narcissistic Personalities² (1970). They were gathered in his first book, Borderline Conditions and Pathological Narcissism³ (1975).

In 1973 Kernberg moved to New York as director of the General Clinical Service of the New York State Psychiatric Institute. In 1974 he became professor of clinical psychiatry at Columbia University and a training and supervising analyst at its Center for Psychoanalytic Training and Research.

In 1976 he joined Cornell as professor of psychiatry and director of the institute for personality disorders at the New York Hospital–Cornell Medical Center. That institute, the Personality Disorders Institute at the Westchester Division of NewYork-Presbyterian Hospital, became the base of his long collaboration with John Clarkin and Frank Yeomans. Together they developed, manualized and tested transference-focused psychotherapy, first set out in Psychotherapy for Borderline Personality⁴ (1999).

Kernberg was president of the International Psychoanalytical Association from 1997 to 2001, and in 2025 he was named its honorary president. His awards include the Heinz Hartmann Award (1972), the Edward A. Strecker Award (1975), the George E. Daniels Merit Award (1981) and the first Sigourney Award. He is a Distinguished Life Fellow of the American Psychiatric Association.

In Borderline Personality Organization¹ Kernberg argued against the view that the patients loosely called borderline occupy an unstable zone between neurosis and psychosis. He held that they share a specific and stable pathological organization of the personality.

Descriptively, he listed symptoms that should raise the suspicion of it. They include chronic diffuse anxiety, polysymptomatic neuroses, polymorphous perverse sexual trends, impulse neuroses and addictions, and lower-level character disorders such as the infantile, narcissistic, as-if and antisocial personalities. No symptom, however, was decisive.

The diagnosis was to be made structurally. Its marks are nonspecific ego weakness; a shift toward primary-process thinking, seen mainly in unstructured psychological tests; a constellation of primitive defences centred on splitting; and a pathology of internalized object relations in which good and bad images of self and others remain unintegrated. Reality testing, by contrast, is largely preserved. He traced the organization to an excess of pregenital, especially oral, aggression, whether constitutional or the result of early frustration, which drives a premature and distorted oedipal development.

The central defence, in his account, is splitting: good and bad images of self and others are actively kept apart. He set it against the repression that dominates neurotic organization. Around it work primitive idealization, projective identification, denial, omnipotence and devaluation (Borderline Conditions and Pathological Narcissism³).

He took Erik Erikson's notion of identity diffusion and made it a structural criterion of borderline organization. He also extended it from a person's concept of themselves to their concept of the significant people in their life.

In Severe Personality Disorders: Psychotherapeutic Strategies⁵ (1984) he condensed structural diagnosis into three criteria: identity integration, the kind of defences used, and the state of reality testing. By these a patient is placed at a neurotic, borderline or psychotic level of personality organization. The structural interview, which he published in 1981 and set out again in that book, is designed to elicit these features. The interviewer clarifies, confronts and interprets what emerges in the encounter itself. The book applied the approach to adolescence and old age and to the classification of personality disorders.

Object Relations Theory and Clinical Psychoanalysis⁶ (1976) gathered papers from 1966 to 1975. In it Kernberg set out what he called a formulation of psychoanalytic object relations theory within ego psychology.

Its starting point was clinical. Borderline patients alternate between contradictory states of mind, each fully conscious and each held apart from the other, and Kernberg held that neither repression, isolation nor denial explains this. He read it as splitting. He made internalized object relations the basic units of the mind: a self-representation and an object representation joined by an affect. Through introjection, identification and ego identity, which he treated as successive levels of internalization, these units form the ego, the superego and the id.

He traced their development in stages. It begins in an undifferentiated state and passes through fused self and object images and then their differentiation. Next, good and bad images are integrated into whole representations, and finally ego and superego are consolidated. Psychosis, borderline pathology and neurosis correspond to successive points on this line. He also argued that affects are the primary building blocks of motivation, and that libido and aggression are organized out of them over the course of development.

The model brought several traditions together. From Edith Jacobson Kernberg took the idea that self-images and object-images, first fused and later differentiated, form the ground of ego and superego. The underlying notion of self and object representations goes back to Heinz Hartmann's distinction of 1950 and Jacobson's development of it. He used Margaret Mahler's account of separation-individuation to locate the fixation points of borderline and narcissistic pathology.

From Melanie Klein he drew his account of the primitive defences. He set it within an ego-psychological developmental frame which, he held, her own theory lacked. From Ronald Fairbairn he took the view of the mind as built of internalized relationships split into separate ego–object units. In an interview he described the synthesis as drawing on the ego-psychological, Kleinian and British middle-group or independent approaches, with an influence from French psychoanalysis added many years later.

The book's second half applied the model. Kernberg proposed a classification of character pathology on three levels, higher, intermediate and lower. Its criteria are the integration of the superego, whether the defences rest on repression or on splitting, and the quality of object relations and identity, and the borderline conditions fall at the lowest level.

Two chapters on love treated the capacity to fall in love and to stay in love as a developmental achievement. It joins sexual desire with tenderness and with a relation to a whole, real person, and narcissistic and borderline pathology limit it. A final chapter extended the theory to small groups and to hospital administration, sketching a psychiatric hospital run as a therapeutic community.

In Factors in the Psychoanalytic Treatment of Narcissistic Personalities² Kernberg described patients whose main disturbance lies in the regulation of self-regard. They show grandiosity alongside feelings of inferiority, an insatiable need for admiration, little empathy and intense envy. Their emotional life is shallow and lacks the capacity for genuine sadness and mourning.

Many of them function socially much better than other patients with severe character pathology. Beneath that surface, he argued, they share the defensive organization of borderline patients. Their distinctive structure is a pathological grandiose self, a fusion of the actual self, the ideal self and the ideal object. It shields them from dependence on others, while devalued images of self and others are projected outward.

He traced this structure to early relations with cold yet overprotective parental figures and to intense oral rage and envy. Pathological narcissism, on this view, differs in kind from normal infantile narcissism. In the transference the grandiose self is activated: the analyst is alternately idealized and devalued, or treated as nonexistent, and is often left bored or reduced to an extension of the patient. Prognosis depends on such factors as the strength of antisocial trends and the capacity to experience depression and guilt.

He set out his differences with Heinz Kohut in Contrasting Viewpoints Regarding the Nature and Psychoanalytic Treatment of Narcissistic Personalities⁷ (1974). The paper was first given at a 1972 panel of the American Psychoanalytic Association and answers The Analysis of the Self⁸ (1971). Kernberg took the clinical picture as broadly agreed and confined the dispute to theory and technique.

Kohut separated narcissistic disorders from borderline states; Kernberg held that most narcissistic personalities rest on a borderline organization, their surface functioning steadied by the grandiose self. Kohut saw pathological narcissism as an arrest at a normal archaic stage; Kernberg saw a specific pathological structure that protects against rage, envy and devalued images of self and others. He also rejected the idea of a line of narcissistic development apart from object love. For him the libidinal investment of self and of objects develop together, and neither can be understood apart from the fate of aggression.

In technique, Kohut advised letting the idealizing and mirror transferences unfold and accepting the patient's idealizing transference as the revival of a normal developmental need. Kernberg read that idealization as defensive. He called for systematic interpretation of the grandiose self as it is enacted in the transference, of the negative transference, and of the envy and devaluation behind it. A treatment that leaves the patient's aggression unanalysed, he warned, may bring adaptation without lasting structural change.

He aligned his own position with Jacobson, Mahler, Joan Riviere, Herbert Rosenfeld and van der Waals. Of Rosenfeld, the author of On the Psychopathology of Narcissism: A Clinical Approach⁹ (1964), Kernberg said that he had taken many aspects of his Kleinian technique while differing on details.

In Severe Personality Disorders⁵, which set his approach to narcissistic personalities against Kohut's, Kernberg also took up the phrase malignant narcissism, used earlier by Erich Fromm. He made it a clinical syndrome combining narcissistic personality, aggression, antisocial traits and paranoid features. The book treats it among the severe regressions of treatment, alongside stalemates, suicide risk, superego pathology and paranoid regression.

From his first book on, Kernberg recommended an expressive psychoanalytic psychotherapy for most borderline patients (Borderline Conditions and Pathological Narcissism³). In it the primitive defences and the negative transference are confronted and interpreted in the here and now. Acting out is limited, and the patient's life is given external structure where needed. For most narcissistic personalities he recommended standard psychoanalysis. Those who function at an overtly borderline level needed the expressive psychotherapy proposed for borderline patients.

He argued for a broad, totalistic view of countertransference: the analyst's whole emotional response to the patient, valuable as information about the primitive transferences of borderline patients. Object Relations Theory and Clinical Psychoanalysis⁶ updated his approach to the transference and countertransference of these patients.

Severe Personality Disorders⁵ presented the expressive psychotherapy, centred on interpreting the transference, together with its adaptation for adolescents and a psychoanalytically based supportive psychotherapy. It set out the indications for each. A final part discussed hospital treatment, including the therapeutic community model and the long-term hospitalization of severe borderline and narcissistic patients.

With Clarkin and Yeomans this approach became transference-focused psychotherapy. Its first manual, Psychotherapy for Borderline Personality⁴, holds that the borderline patient's inner world is split into polarized, affect-laden pairs of self and other representations. These pairs, idealized and persecutory, victim and destroyer, are enacted in the relationship with the therapist.

The treatment is held twice weekly and begins with a negotiated contract. The contract sets the frame, the responsibilities of patient and therapist, and how threats to the treatment will be handled. The work is organized as long-range strategies, session-by-session tactics and moment-to-moment techniques. Interventions follow a hierarchy of priorities: threats to life and to the continuation of treatment come first, then acting out that undermines the work, and only then the themes of the transference.

The therapist identifies the dominant dyad in the here and now and clarifies, confronts and interprets it from a stance of technical neutrality, using the countertransference as information. The aim is structural change: integrating split representations so that the patient can experience self and others as whole and complex.

In an interview Kernberg described a randomized trial at his institute. It assigned three groups of 40 patients with borderline personality disorder to transference-focused psychotherapy, dialectical behaviour therapy and a supportive therapy. Its purpose, he said, was not a horse race between treatments but a study of which processes in treatment are linked to which mechanisms of change.

In the same interview he criticized the lack of systematic empirical research in psychoanalysis. He argued that long-term psychotherapies should be manualized and tested scientifically.

The 1967 paper became the founding statement of Kernberg's structural approach to borderline and narcissistic pathology. It joined Klein's ideas on splitting and projective identification to Jacobson's ego psychology of self and object representations (Borderline Personality Organization¹). Severe Personality Disorders⁵ became a standard clinical reference, and its structural interview the basis of later assessment instruments.

The first manual of transference-focused psychotherapy underlay the randomized trials of the method that followed (Psychotherapy for Borderline Personality⁴). It was translated into German in 2001 and revised in 2006 as Psychotherapy for Borderline Personality: Focusing on Object Relations. Randomized controlled trials of the treatment have found it effective in reducing suicidality, anger and impulsivity, and in improving reflective functioning and interpersonal capacity. Among them is a trial by Doering and colleagues (2010) that compared it with treatment by community psychotherapists.

His disagreement with Kohut set an emphasis on aggression, primitive defences and the interpretation of defensive idealization against an emphasis on developmental arrest and the acceptance of the patient's idealizing needs. The two positions shaped one of the central debates of late twentieth-century psychoanalysis (Contrasting Viewpoints⁷). Kernberg has said that, having been accused of being concerned only with hatred and aggression, he thought it would be fun to write about love, and he did so in Love Relations: Normality and Pathology.

Kernberg's early papers in the Journal of the American Psychoanalytic Association include Borderline Personality Organization¹ (1967), Factors in the Psychoanalytic Treatment of Narcissistic Personalities² (1970) and Contrasting Viewpoints Regarding the Nature and Psychoanalytic Treatment of Narcissistic Personalities: A Preliminary Communication⁷ (1974).

His books include Borderline Conditions and Pathological Narcissism³ (1975), Object Relations Theory and Clinical Psychoanalysis⁶ (1976), Severe Personality Disorders: Psychotherapeutic Strategies⁵ (1984), Aggression in Personality Disorders and Perversions (1992) and Love Relations: Normality and Pathology.

With Clarkin and Yeomans he wrote Psychotherapy for Borderline Personality⁴ (1999), revised in 2006 as Psychotherapy for Borderline Personality: Focusing on Object Relations. With colleagues he also wrote Handbook of Dynamic Psychotherapy for Higher Level Personality Pathology (2007), with Eve Caligor as lead author, and Transference-Focused Psychotherapy for Borderline Personality Disorder: A Clinical Guide (2015). He wrote Treating Pathological Narcissism with Transference-Focused Psychotherapy (2021) with Diana Diamond, Frank Yeomans and Barry Stern.