The Borderline Patient: An Overview

by William Goldstein, M.D.

Who may benefit from reading this article: Psychiatrists, primary care physicians, psychologists, social workers, psychiatric nurses and mental health care practitioners.

Educational Objectives

After reading this article, you will be familiar with:

The term borderline first appeared in the psychoanalytic literature only later to become a term generally used among psychiatrists and other mental health professionals. Early on, analysts were confronted with the situation in which a patient came for analysis-seemed analyzable-then when on the couch proved to have great difficulties with the process. Obviously, the patient was much more disturbed than originally thought.
As analysts had negative experiences with these types of patients, they wrote papers using different terms to describe these more troubled individuals. The early psychoanalytic papers on the borderline patient fall into two groups. One group views these patients as having a mild form of schizophrenia, whereas a second group views them as a distinct and separate group of individuals, neither neurotic nor psychotic, operating psychopathologically on a level between the two.

Many of the patients from both of these groups would be considered borderline today. Consequently, in the first group we have what Zilboorg (1941) described as the ambulatory schizophrenic, what Hoch and Polatin (1949) described as the pseudoneurotic schizophrenic, and what Bychowsky (1953) described as the latent psychotic. In the second group, we have Stern's (1938) borderline patient, Deutsch's (1942) as-if personality, Knight's (1953) borderline and Frosch's (1964) psychotic character.

In 1967, Otto Kernberg, M.D., wrote his seminal paper providing for many clinicians an integration and synthesis of all earlier writings on the borderline patient, offering a unified and comprehensive framework for description, definition and understanding. Although he elaborated, updated and modified his original article, contributing many additional ideas regarding the borderline patient (1984, 1985ab, 1989), his core ideas are still best expressed in his 1967 paper.

For some, Kernberg's writings, characterized by terminology and language related to an underlying sophisticated psychoanalytic object relations theory, are difficult and confusing. Despite this, his work remains central to current psychoanalytic thinking regarding the borderline patient.

With this in mind, I have presented Kernberg's work in a more simplified fashion (Goldstein 1985, 1996). I've used the term ego-psychological diagnostic approach to describe a way of dynamically presenting the borderline patient based on a profile of ego strengths and ego weaknesses. This ego-psychological diagnostic approach is based on Kernberg's ideas, but it simplifies and sometimes modifies his work, and additionally includes aspects of the borderline patient not stressed by him. This approach provides a current summary of the psychodynamic concept of the borderline individual.

It should also be noted that although the ego-psychological diagnostic approach, with its psychodynamic underpinnings, is very different in theoretical focus from the Diagnostic and Statistical Manual (DSM-IV), it is clearly compatible with this most commonly used diagnostic system. All items on the DSM-IV checklist of borderline personality traits and symptoms correspond to the list of ego strengths and weaknesses of the ego-psychological diagnostic approach. DSM-IV is a more descriptive and atheoretical diagnostic system, more conducive to psychological research. The ego-psychological diagnostic approach, by contrast, is more conducive to understanding and psychotherapeutic intervention.

Diagnostic Approach

The borderline patient is pictured as having a specific stable pathological personality organization, characterized by a specific kind of underlying structural configuration. It should be noted that the stability here refers to the structural configuration, and not to mood and behavioral fluctuations. The structural configuration is viewed as including a distinct kind of ego and superego functioning, and a distinct pattern of instinctual drive organization. This configuration is quite stable and resistant to change except through intensive psychotherapeutic intervention or with the passage of much time.

The borderline patient is viewed as having one of four structural configurations, within which all patients can be classified diagnostically. Thus, the borderline structural configuration, along with the normal-neurotic, the narcissistic, and the psychotic, provide first-level "structural" diagnoses under which all other descriptive diagnoses fall. Although the structural configuration includes a specific kind of ego and superego functioning, as well as a specific pattern of instinctual drive organization, the focus for diagnostic purposes will be largely on the ego.

Relying on the model of Beres (1956) and Bellak (1958), the borderline patient's ego will be viewed in accordance with the various ego functions. The borderline individual is thus described as having a particular ego structure, consisting of a specific pattern of relative ego strengths and underlying ego weaknesses.

It must be stressed that these four strengths are only relative; they easily break down to various degrees in various situations. Because these four relative strengths stand out superficially, they enable the borderline individual to present a fairly "normal" appearance. These relative strengths, particularly the first two, most clearly differentiate the borderline from the more psychotic individual.

In contrast to the strengths, which stand out on a superficial level, these weaknesses only become clearly apparent with in-depth understanding. Except during regressed states, a detailed history or a relationship over time is needed for these weaknesses to clearly emerge. Because these weaknesses are beneath the surface and are not detected superficially, they do not detract from the borderline's appearance of normality. These underlying weaknesses, however, most clearly differentiate the borderline from the more neurotic individual. A brief look at the relative ego strengths and underlying ego weaknesses will now be undertaken.
First, we will look at the relative ego strengths.

Reality Testing

The strength here is that on a surface level, and in day-to-day functioning, reality testing is basically intact. The weakness can emerge under stress and in very close interpersonal situations where there is a tendency for this ego function to regress, sometimes leading to brief psychotic episodes. These episodes are short-lived, spontaneously reversible and always related to clear-cut precipitating events. These transient psychotic episodes under stress are allowable, but certainly not mandatory to making the diagnosis of borderline.

Thought Processes

The strength is that in day-to-day functioning and in structured situations, thought processes are predominantly secondary process. The weakness can come about under stress and in unstructured situations (such as projective psychological testing) when primary process sometimes emerges. There is the tendency to find a psychological test pattern of secondary process on the Wechsler Adult Intelligence Scale and primary process on the Rorschach test.

Interpersonal Relations

The strength here is that the borderline patient often seems to do adequately in interpersonal relations. On the surface the patient seems to "relate" to others, can have many acquaintances and sometimes can maintain long-term relationships. Weaknesses emerge when, under closer scrutiny, it becomes apparent that the relationships are often characterized by a striking lack of depth and a lack of concern for the other individual. The other person is seen as someone who can be used to meet the borderline patient's needs rather than as a person in his or her own right. Empathy is lacking, and the borderline individual often vacillates between superficial relationships and intense, dependent relationships that are marred by primitive defenses.

Adaptation to Reality

The strength here is that adaptation to reality is often superficially intact. The borderline patient may seem of normal appearance and may seem to display adequate achievement in work or school. Weaknesses in this area emerge when, under closer scrutiny, there is often a feeling that the adaptation is far from optimal. There are certain "exceptional" borderline patients who can maximize certain strengths and adapt adequately over time, particularly in structured settings. These individuals often do quite well professionally, while typically displaying much more chaos in their social lives. Typically, these people display certain marked ego strengths together with their weaknesses. Strengths often include high intelligence and the ability to use obsessive-compulsive defenses.

We will now turn to the underlying ego weaknesses.

Impulse Control and Tolerance

Invariably, the borderline patient displays the combination of poor frustration tolerance and poor impulse control. There is an inability to delay, a demand for immediate gratification and a proclivity to act out under stress. To make matters more difficult, these characteristics are not infrequently combined with a sense of entitlement, especially in patients with narcissistic features. These difficulties most frequently present themselves clinically as a tendency for states of disruptive anger; use of drugs and alcohol to avoid frustration and obtain temporary gratification; and an inclination to flee the work or interpersonal situation under stress.

Primitive Ego Defenses

Emphasis here is on the proclivity to use primitive ego defenses. The borderline patient does not use primitive defenses all the time. In day-to-day functioning, the patient tends to use primitive defenses somewhat more than the neurotic, but under stress tends to rely on these defenses. Following Vaillant (1971), one can classify defenses into a hierarchy of five levels of increasing psychopathology: mature, neurotic, immature, borderline and psychotic. Actually, Vaillant conspicuously leaves out the borderline defenses from his classification. Nevertheless, the borderline defenses can be conveniently added to his classification, between his immature and psychotic levels. With this hierarchy in mind, these generalizations can be made. In day-to-day functioning, the borderline patient, with marked individual variation, uses a combination of mature, neurotic, immature and borderline defenses. Under stress, he or she displays a marked tendency to rely on the borderline defenses. In marked regressions, the patient may also use some psychotic defenses.

Borderline defenses include splitting, primitive idealization, projection, projective identification, primitive denial, omnipotence and devaluation. These are the borderline patient's defenses considered by Kernberg to be pathognomonic. In addition to the above, primitive defenses are thought to include acting out and psychotic defenses.

Identity Diffusion

This term refers to an identity that is not integrated or cohesive, but diffuse. It is an identity based on multiple contradictory unintegrated self images. Correspondingly, there are multiple contradictory unintegrated object images. At one time one self image is evoked, at another time, a different one. This same applies to object images. Neither a comprehensive view of the self nor of objects has ever been attained. Sometimes the borderline patient experiences this problem as an inner lack or void, a sense of emptiness or depletion.

Affective Instability

Affective instability can include the presence of irritability; intense affect, usually depressive or hostile; anger as main affect experienced; and depressed, lonely and empty feelings. Aggression is not utilized in constructive, ego-syntonic, adaptive ways, such as sublimations, work, recreation and enjoyment. Instead, it often breaks through directly or is defended against and results in other ego-dystonic affect states, such as depression, boredom and emptiness. Often there are rapid and dramatic swings from one affect state to another. These swings occur on a daily basis in response to environmental factors, in contrast to the more sustained mood shifts typical of the bipolar patient.

In using the ego-psychological diagnostic approach, one does not need to demonstrate all four ego strengths and all four ego weaknesses to make the diagnosis. In contrast, one looks at the overall pattern of ego strengths and weaknesses and establishes into which large grouping (the normal-neurotic, the narcissistic, the borderline or the psychotic) the patient best fits. The borderline concept outlined here is quite broad. It can be thought of as including a heterogeneous group of patients with varying degrees of pathological functioning, various personality styles and various symptoms.

What differentiates this heterogeneous group from the other large groupings is the underlying pattern of ego strengths and ego weaknesses just elaborated. The ego-psychological diagnostic approach is thought to complement the work of Meissner (1984, 1988) regarding the borderline spectrum. I agree with Meissner, who views the borderline conditions as consisting of a heterogeneous group of patients represented by a spectrum of levels and degrees of pathological personality functioning.


Turning to developmental and etiological theory, I will briefly compare what I find to be the two most helpful psychodynamic models regarding the borderline patient (Goldstein 1989). I am referring to the theories of Kernberg (1967, 1984, 1985a) and Adler (1985). According to Kernberg, the failure to attain libidinal object constancy, with the concomitant difficulty in integrating positive and negative introjects, is the core problem of borderline patients. The presence of an excessive amount of aggressive drive in the earliest years of life is the basic etiologic factor. This excessive aggressive drive interferes with the unfolding of the separation-individuation process in such a way that splitting is reinforced and libidinal object constancy is never truly attained. According to Kernberg, this excessive aggressive drive either is due to congenital factors or is secondary to severe early frustration.

To elaborate briefly, Kernberg believes that the infant organizes experiences according to whether the experiences are perceived as positive or negative. Thus, the infant tends to view objects as either all good or all bad. This tendency to perceive objects as all good or all bad is called splitting. It is not until the child is between the ages of 2 and 3 (usually closer to 3) that the child stops using splitting as a primary organizing principle and develops libidinal object constancy. Libidinal object constancy (McDevitt, 1975) is defined as the ability of the child to maintain an emotional image of the mother as being basically good but as having both good and bad qualities, an emotional image that changes little under frustration or during a mother's absences.

Although Adler agrees with Kernberg that borderline patients experience difficulty in integrating good and bad self images and object images, he believes that this problem arises at a later point in development and at a later time in treatment than generally thought. He does not consider this difficulty the core problem in borderline patients. For Adler, the core difficulty is a functional insufficiency and instability of a certain kind of introject, which he calls the holding introject.

Holding introjects allows people to provide holding and soothing for themselves without the need for positive feedback from others. It is the lack of this specific holding introject, rather than a difficulty in integrating contradictory introjects, that Adler considers crucial. According to Adler, borderline individuals are unable to adequately internalize holding introjects; they remain dependent on other people to provide needed holding and soothing.

Although some (Waldinger, 1993) have focused on the differences in the theories of Kernberg and Adler, I emphasize their compatibility. I think that the two can be combined into one integrated model, emphasizing two core problems for the borderline individual: (1) the deficiency in holding (positive) introjects, and (2) the difficulty in integrating positive and negative introjects.

This is a good point to comment on environmental etiology. Deprivation, privation and neglect have been implicated as etiological in some borderline patients. In others, a pattern of overindulgent, overintrusive and overstimulating mothering has been noted. Here the primary caretaker forms an engulfing, suffocating and fused relationship with the borderline-to-be child, not allowing the child to separate and individuate and to become autonomous (Masterson, 1976, 1978; Rinsley, 1978, 1982). Additionally, there is a growing literature showing a high percentage of borderline individuals with a history of trauma, of physical and/or sexual abuse (Herman et al., 1989; Zanarini et al.,1989; Ogata et al.,1990; Stone, 1990). Neglect, overindulgence and trauma all can play an etiologic role, but in different borderline patients.

Without going into the details, the significance of the work of Masterson (1976, 1978), Rinsley (1978, 1982), Giovacchini (1983, 1993) and Gunderson (1984) should be noted. Masterson and Rinsley are associated with the overindulgent mothering theory, Giovacchini has his own highly elaborate and complex but clearly relevant formulations, and Gunderson offers a very useful integrative and practical perspective.


Turning to prognosis, the studies of McGlashan (1985, 1986) and Stone (1990) provide important and similar findings. Stone's book, describing the long-term follow-up study of 502 hospitalized patients, will be discussed here. The findings of this study certainly create an optimistic attitude regarding prognosis.

To get to the bottom line first, the borderline patients in the study, given time, tended to get better. In fact, two out of three of these patients got better; this was in marked contrast to the group of schizophrenic patients, in which approximately one out of 10 improved.

Despite this overall positive picture, outcome varied considerably, in accordance with a variety of factors. The best outcome was associated with patients having artistic talent, an obsessive-compulsive personality style, a very high IQ and/or general attractiveness. The general characteristics of likableness, candor, perseverance and talent were also associated with good outcome. While a family history of alcohol/substance abuse and current alcohol/substance abuse is correlated with a poor prognosis in borderline disorder (McGlashan, 1986), Stone reported that definitive treatment in Alcoholics Anonymous, was associated with a good prognosis.

Poor outcome was associated with patients who had served time in jail, had committed rape and who had an antisocial personality. Males who eloped, women who were subjected to incest by their fathers and patients suffering brutality in childhood had negative outcomes. Fifty percent of the borderline women in the study had a history of incest, as compared to 5% of women in general. Incest was most pathological when it was transgenerational, involved force and was chronic. The combination of incest plus parental brutality was seen to be devastating.

The borderline concept remains a difficult and confusing topic to many. Focus on the ego-psychological diagnostic approach, together with an awareness of the developmental and etiological theories, plus outcome studies can contribute to making this concept more understandable.

Dr. Goldstein is a teaching analyst and director of the Adult Psychotherapy Training Program at the Baltimore-Washington Institute for Psychoanalysis, as well as clinical professor of psychiatry at the Georgetown University Medical Center. He has written extensively in professional journals and has published three books: An Introduction to the Borderline Conditions (Jason Aronson 1985), Dynamic Psychotherapy with the Borderline Patient (Jason Aronson 1996), and A Primer for Beginning Psychotherapy (Brunner/Mazel 1998).


Adler G (1985), Borderline Psychopathology and Its Treatment. New York: Jason Aronson.

American Psychiatric Association (1994), Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington: American Psychiatric Association.

Bellak L (1958), Schizophrenia: A Review of the Syndrome. New York: Logos Press.

Beres D (1956) Ego deviation and the concept of schizophrenia. Psychoanal Study Child 2:164-235.

Bychowsky G (1953), The problem of latent psychosis. J Am Psychoanal Assoc 4:484-503.

Deutsch H (1942), Some forms of emotional disturbance and their relation to schizophrenia. Psychoanal Q 11:301-321.

Frosch J (1964), The psychotic character: clinical psychiatric considerations. Psychiatr Q 38:81-96.

Giovacchini P (1993), Borderline Patients, the Psychosomatic Focus and the Therapeutic Process. Northvale, N.J.: Jason Aronson.

Giovacchini P (1983), Psychoanalysis of Character Disorders and the Study of Ego Adaptation. New York: Jason Aronson.

Goldstein WN (1989), An integration of the views of Kernberg and Adler on borderline pathology. Bull Menninger Clin 53(4):300-309.

Goldstein WN (1985), An Introduction to the Borderline Conditions. Northvale, N.J.: Jason Aronson. Goldstein WN (1996), Dynamic Psychotherapy with the Borderline Patient. Northvale, N.J.: Jason Aronson.

Goldstein WN (1998), A Primer for Beginning Psychotherapy. New York: Brunner/Mazel.

Gunderson JG (1984), Borderline Personality Disorder. Washington: American Psychiatric Press.

Herman JL, Perry JC, van der Kolk BA (1989), Childhood trauma in borderline personality disorder. Am J Psychiatry 146(4):490-495. See comments.

Hoch PH, Polatin P (1949), Pseudoneurotic forms of schizophrenia. Psychiatr Q 23:248-276.

Kernberg OF (1967), Borderline personality organization. J Am Psychoanal Assoc 15(3):641-685.

Kernberg OF (1984), Severe Personality Disorders: Psychotherapeutic Strategies. New Haven: Yale University Press.

Kernberg OF (1985a), Borderline Conditions and Pathological Narcissism. New York: Jason Aronson.

Kernberg OF (1985b), Internal World and External Reality: Object Relations Theory Applied. New York: Jason Aronson.

Kernberg OF, Seltzer M, Koenigsberg H, Carr A et al. (1989), Psychodynamic Psychotherapy of Borderline Patients. New York: Basic Books.

Knight R (1953), Borderline states. Bull Menninger Clin 17:1-12.

Masterson J (1978), New Perspectives on Psychotherapy of the Borderline Adult. New York: Brunner/Mazel.

Masterson J (1976), Psychotherapy of the Borderline Adult: A Developmental Approach. New York: Brunner/Mazel.

McDevitt JB (1975), Separation-individuation and object constancy. J Am Psychoanal Assoc 23(4):713-742.

McGlashan TH (1985), The prediction of outcome in borderline personality disorder: Part V of the Chestnut Lodge Follow-up Study. In: McGlashan T, ed. The Borderline: Current Empirical Research. Washington: American Psychiatric Press.

McGlashan TH (1986), The Chestnut Lodge follow-up study: III. Long-term outcome of borderline personalities. Arch Gen Psychiatry 43(1):20-30.

Meissner WW (1984), The Borderline Spectrum: Differential Diagnosis and Developmental Issues. New York: Jason Aronson.

Meissner WW (1988), Treatment of Patients in the Borderline Spectrum. Northvale, N.J.: Jason Aronson. Ogata SN, Silk KR, Goodrich S et al. (1990), Childhood sexual and physical abuse in adult patients with borderline personality disorder. Am J Psychiatry 147(8):1008-1013. See comments.

Rinsley DB (1982), Borderline and other Self Disorders: A Developmental and Object Relations Perspective. New York: Jason Aronson.

Rinsley DB (1978), Borderline psychopathology: a review of etiology, dynamics and treatment. Int Rev Psychoanal 5:45.

Stern A (1938), Psychoanalytic investigation and therapy in the borderline group of neurosis. Psychoanal Q 7:467-489.

Stone MH (1990), The Fate of Borderline Patients: Successful Outcome and Psychiatric Practice. New York: Guilford Press.

Vaillant GE (1971), Theoretical hierarchy of adaptive ego mechanisms: a 30-year follow-up of 30 men selected for psychological health. Arch Gen Psychiatry 24(2):107-118.

Waldinger RJ (1993), The role of psychodynamic concepts in the diagnosis of borderline personality disorder. Harvard Review of Psychiatry 1:158-167.

Zanarini MC, Gunderson JG, Marino MF et al (1989), Child experiences of borderline patients. Compr Psychiatry30(1):18-25.

Zilboorg G (1941) Ambulatory schizophrenias. Psychiatry 4:149-155.