GeistHaus

Personality / Organization

Borderline Level of Organization

The borderline level of personality organization is the broad middle region of the severity spectrum, between the neurotic and psychotic levels. The term grew out of decades of clinical experience with people who seemed, in McWilliams’s phrase, “too sane to be considered crazy, and too crazy to be considered sane” (McWilliams, 2011, p. 52). Following Otto Kernberg, the PDM-3 defines it by three features: identity diffusion (a sense of self that is not well integrated), a predominance of primitive defenses, and reality testing that is fragile but intact (McWilliams & Shedler, 2026, p. 655). Clinicians have long described its quality as a “stable instability” (McWilliams, 2011, p. 34).

The word “borderline” is used in three different ways, and confusing them causes real harm. On this site, borderline level means a degree of personality integration that can underlie any personality style. Borderline personality disorder is a specific diagnosis in the DSM. The emotionally dysregulated style is the PDM-3’s name for the personality pattern the DSM diagnosis describes. Everyone with that style functions at the borderline level, but many people at the borderline level do not have it (McWilliams & Shedler, 2026, p. 696). The Boundaries section below spells out the differences.

This level covers a lot of people. As a general rule, the PDM-3 says, people whose personality difficulties are severe enough to warrant any DSM personality disorder diagnosis function here (p. 658). Many are bright, talented, and articulate (McWilliams, 2011, p. 86); they are not always unpleasant people, despite their reputation among professionals, and they usually respond to good treatment (p. 89). Their suffering is often intense, and its causes, the research suggests, are many: constitutional vulnerability, misattuned early attachment, and trauma all appear to play a part (p. 54).

Hallmarks

Identity diffusion
The PDM-3 describes an inability to integrate different aspects of the self into a coherent whole. Attitudes, values, goals, and feelings about oneself tend to shift, and self-perceptions can swing between contradictory extremes; people may look quite different on different occasions (McWilliams & Shedler, 2026, p. 659). Asked to describe important people in their lives, they often give global, flat answers rather than three-dimensional portraits (McWilliams, 2011, p. 64). Unlike people at the psychotic level, though, as McWilliams puts it, “they know they exist” (p. 64). The relevant M-axis capacity is differentiation and integration.
Primitive defenses
The PDM-3 names splitting (of the self-image and of others’ image) into all-good and all-bad, projection, and projective identification as the most noted, along with denial, withdrawal, wholesale introjection, omnipotent control, acting out, severe dissociation, and extreme idealization and devaluation. Vaillant’s research adds hypochondriacal concerns and passive aggression, and somatization is common (McWilliams & Shedler, 2026, pp. 658–660). A distinctive feature of projection at this level is its certainty: the person regards it as fact that the projected feelings belong to the other person (p. 659). The PDM-3 suggests that what most separates this level from the neurotic one may be the relative absence of mature defenses under stress, rather than the mere presence of costly ones (p. 658). See defensive functioning.
Reality testing
Intact but fragile. Interviewed thoughtfully, people at this level show an appreciation of reality however dramatic their difficulties look; asked whether others might find something about them unusual, they can see the point (McWilliams, 2011, pp. 64–65). In intense relationships, including therapy, they may temporarily lose this, for instance becoming convinced that a therapist is exactly like a parent (p. 52; McWilliams & Shedler, 2026, p. 655).
Object relations
Relationships tend to swing between closeness that feels engulfing and distance that feels like abandonment (McWilliams, 2011, p. 66). Others are often experienced as all good or all bad, and the cast may change while the roles stay the same (McWilliams, 1999, p. 77). The central anxiety is separation anxiety (McWilliams, 2011, p. 55). See trust, empathy, and intimacy.
Other capacities
The PDM-3 calls emotional dysregulation the most salient observable feature of the borderline range: overwhelming, rapidly shifting feelings, chaotic relationships, and difficulties with work and with impulse regulation (McWilliams & Shedler, 2026, p. 658). In periods of distress, typically set off by a threat to an important attachment, some people at this level are at risk of self-destructive behavior, which the PDM-3 understands as a desperate effort to regulate unbearable emotion, a failure of impulse control, or both (p. 658). This picture is truest of the more dramatic presentations; some people at this level are quiet and withdrawn (p. 656). Following Peter Fonagy, McWilliams notes that people at this level often struggle to mentalize, to find meaning in their own behavior and others’ in terms of separate minds (2011, p. 64), and that they have trouble with affect tolerance and regulation, going quickly to anger where others might feel shame, envy, or sadness (p. 64). On the M axis, borderline functioning spans three bands, M4 to M6, from moderate impairments to significant defects (Lingiardi, Muzi, & Bornstein, 2026, pp. 738–740).

How It Feels From the Inside

The experience of self at this level is likely to be full of inconsistency and discontinuity (McWilliams, 2011, p. 64). A person may feel fine one day and, the next, have no access to any good feeling about themselves at all; the PDM-3 describes people who, when feeling good, seem untroubled by how bad things recently were, and who, when low, seem unable to reach any positive self-perception (McWilliams & Shedler, 2026, p. 659). McWilliams describes a woman who flew into a fury at an intake form asking her to complete the sentence “I am the kind of person who…”; years later she could fill it in and wondered why it had enraged her (2011, p. 64).

Relationships carry a painful double bind. Closeness brings fear of being engulfed and controlled; being alone brings a sense of being traumatically abandoned. McWilliams, drawing on James Masterson, describes people caught going back and forth, unable to rest comfortably at either distance (2011, p. 66). When the person to whom they are attached is absent, they may feel empty or insubstantial (McWilliams, 1999, p. 79).

From the inside, the problem rarely looks like personality. McWilliams notes that people at this level usually come for help with panic, depression, or stress-related illness, or at someone else’s urging, and having never had any other kind of character, they have little emotional basis for knowing how it would feel to be different. “They just want to stop hurting” (2011, pp. 65–66). Well-meant comments can land as attacks, because in an upset state there is not yet an observing part of the self able to take them as information (p. 66).

Across the Styles

“Borderline,” as a level, encompasses great diversity. McWilliams stresses that a depressive person at this level is quite different from a narcissistic, hysterical, or paranoid one (2011, p. 83). Some patterns from the PDM-3:

In Therapy

Two terms help here. Transference is the way a person experiences the therapist through the lens of earlier relationships; countertransference is the therapist’s emotional response. At this level transferences tend to be intense and hard to recognize as transferences. The PDM-3 gives the example of a man who, told he may be reacting to his therapist as he did to his father, concludes that it is terrible the therapist is exactly like his father (McWilliams & Shedler, 2026, p. 663). Therapists, in turn, often feel strong and upsetting reactions: rescue fantasies, punitiveness, helplessness, or confusion. In hospitals, staff can split into camps that see the same patient as deprived and needing love, or as manipulative and needing limits (McWilliams, 2011, p. 67; McWilliams & Shedler, 2026, p. 660). The PDM-3 adds that people at this level often exert strong pressure on therapists to act, sometimes in ways at odds with their professional role, which helps explain some boundary violations (p. 660).

Despite different theories, the psychodynamic, cognitive-behavioral, and other treatments for this level show what McWilliams calls remarkable practical consensus (2011, p. 84). Among the shared principles:

Research-tested approaches include transference-focused psychotherapy, mentalization-based treatment, dialectical behavior therapy, schema therapy, and good psychiatric management, among others (McWilliams & Shedler, 2026, p. 661). Change takes time. McWilliams notes that building a working alliance can take years, though the most disabling behaviors may ease well before then (2011, pp. 66, 84), and that crises often come precisely when a person first begins to feel hope (p. 18; McWilliams & Shedler, 2026, p. 663). Where a person sits within the borderline range shapes the approach: more exploratory work toward the neurotic border, more supportive, capacity-building work toward the psychotic border (McWilliams, 2011, p. 83; McWilliams & Shedler, 2026, p. 658).

How Different Thinkers See It

Kernberg

Otto Kernberg gave the level its structural definition: identity diffusion, a predominance of primitive defenses, and generally intact reality testing (Kernberg, 1967, 1975; McWilliams & Shedler, 2026, p. 655). He proposed that adequacy of reality testing, rather than “insight into illness,” separates borderline from psychotic organization (McWilliams, 2011, p. 65). In his view, split-apart, unintegrated images of self and others underlie both the identity disturbance and the troubled relationships (Blüml & Doering, 2021). His transference-focused psychotherapy uses clarification, confrontation, and interpretation of what happens in the relationship with the therapist to change this underlying impairment (Hörz-Sagstetter, Ohse, & Kampe, 2021). After the DSM adopted the term, Kernberg, one of the originators of the concept, began distinguishing borderline personality organization from DSM borderline personality disorder (McWilliams, 2011, p. 53).

Mahler and Masterson

Margaret Mahler described a “rapprochement” subphase, around age two, when a child who has gained some independence still needs reassurance that a caregiver remains available. James Masterson saw borderline difficulties as rooted in this phase, with its back-and-forth between “I can do it myself” and collapse into need, and he described the “abandonment depression” people feel when alone (McWilliams, 2011, pp. 66–67, 91). McWilliams notes that whether or not his ideas about cause are right, his observations about entrapment in dilemmas of separation help make sense of these struggles (p. 67).

Erikson and Klein

In Erikson’s terms the central struggle is autonomy versus shame and doubt (McWilliams, 2011, p. 53; McWilliams, 1999, p. 77). Melanie Klein’s account of splitting, keeping good and bad experiences of the same person apart, underlies the PDM-3’s description of borderline defenses (McWilliams & Shedler, 2026, pp. 658–659).

Kohut and self psychology

Where Kernberg saw serious narcissistic pathology as structural, Heinz Kohut saw stunted normal development; Kernberg in turn suggested that Kohut’s empathic approach was a form of supportive therapy suited to narcissistic patients in the borderline-to-psychotic range, even though Kohut worked mostly with high-functioning people (McWilliams, 2011, pp. 188–189). Self psychology frames these difficulties around the cohesion or fragmentation of the self. Lawrence Hedges, whose scheme McWilliams summarizes, suggested listening for self-psychological themes of self-cohesion and fragmentation with narcissistically organized people, but for object-relational themes of merger versus abandonment, affect differentiation, and separation with people organized at the borderline level (McWilliams, 2004, p. 138).

Fonagy and the mentalization tradition

Peter Fonagy and colleagues describe borderline difficulties as rooted in insecure attachment and a failure of the reflective function that finds meaning in one’s own and others’ behavior (McWilliams, 2011, p. 64). Their mentalization-based treatment aims to strengthen that capacity (p. 71).

Winnicott, Searles, and the use of feelings

D. W. Winnicott’s paper “Hate in the Countertransference” (1949/1958) and Harold Searles’s frank accounts of the therapist’s emotional storms helped clinicians see their own intense reactions to deeply troubled patients as a source of understanding rather than only a failing (McWilliams, 2011, p. 34).

Where they disagree

Theories of cause have diverged widely: constitutional vulnerability, failures in the separation phase, early parent–child interaction, family systems, social conditions, attachment, and trauma have all been proposed (McWilliams, 2011, p. 54). The PDM-3 notes twin studies attributing about 40% of the variance in borderline traits to genetics (p. 656). McWilliams concludes that borderline psychology is probably not a single entity but multidetermined (2011, p. 54).

In Other Systems

These correspondences are approximate; the first two are this site’s mapping.

Examples

Donna

When McWilliams first met Donna in the early 1970s, she was a 23-year-old college student who had been hospitalized, heavily medicated, and diagnosed with schizophrenia. She was hostile, desperate, deeply suspicious, and self-destructive, and her anxiety in interviews disorganized her so much that she sounded flagrantly psychotic. In retrospect McWilliams sees her as organized at a borderline level, at the border of the psychoses rather than the border of the neuroses, since she never had fully elaborated hallucinations or delusions. Over a very long therapy Donna made substantial and lasting progress. Decades later she still had anxieties, but she could use other people as supports, felt proud of her daughter and her growth, and agreed to the publication of her story in the hope that it would encourage therapists to keep faith with their most difficult clients.

Nancy McWilliams, Psychoanalytic Psychotherapy (2004), pp. 219–222, 237–238

The long silence

A 22-year-old man in therapy with McWilliams began falling silent for 15 or 20 minutes at a time. She sensed that if she stayed quiet he would feel neglected, and if she spoke he would feel taken over, a mirror of his fear of both abandonment and engulfment. She asked him how he wanted her to respond. He first asked her to draw him out, then, when she noted she could not read his mind, decided she should wait. He came to three sessions in which he said nothing at all, and she was at peace with it. A couple of years later he told her that her willingness to take his direction marked the beginning of his ability to feel like a separate person in someone else’s presence.

Nancy McWilliams, Psychoanalytic Diagnosis (2nd ed., 2011), pp. 90–91

Kai

In one of the PDM-3’s online case illustrations, Kai, a nonbinary 27-year-old, struggled with long-standing relational difficulties, little trust in intimacy, anxiety, social withdrawal, and binge eating. The authors rate Kai’s organization as borderline/neurotic, with a poor sense of identity but good reality testing and impulse control, close to what this page calls the higher borderline range. They warn explicitly that a traditional reading could have pathologized Kai’s nonbinary identity itself. Over therapy with a clinician who had to work through his own discomfort, Kai developed a more integrated sense of self, greater resilience, a partner, and a circle of trusted friends.

Lingiardi & McWilliams (Eds.), PDM-3 Online Supplement (2026), pp. 55–57

Boundaries

Borderline level vs. borderline personality disorder
The borderline level is a degree of personality integration that can underlie any style: depressive, schizoid, narcissistic, paranoid, and so on. Borderline personality disorder is a DSM category. When it entered DSM-III in 1980, McWilliams argues, the concept lost its original meaning as a level of functioning, because the DSM definition drew on research with people most analysts would have seen as having a histrionic style at the borderline level (2011, pp. 52–53). The PDM-3 calls the DSM diagnosis “a conflation of borderline level of personality organization and histrionic personality style” (McWilliams & Shedler, 2026, p. 655). So everyone who meets criteria for the DSM diagnosis functions at the borderline level, but many people at the borderline level do not have the style that diagnosis describes (p. 696). ICD-11’s “borderline pattern” qualifier follows the DSM meaning (Bach & First, 2018).
Borderline level vs. the emotionally dysregulated style
To reduce confusion, PDM-3 keeps “borderline” only for the level and calls the personality style described by the DSM diagnosis the emotionally dysregulated style (McWilliams & Shedler, 2026, p. 656). This style is the one style tied to a single level: all people with it function at the borderline level, but not all people at the borderline level have it (p. 696). A quiet, withdrawn person with a schizoid style at the borderline level is a clear example of the level without the style (p. 656).
Higher vs. lower borderline
The PDM-3 divides the level into higher ranges, toward the border with the neuroses, describing people whose overall presentation is closer to neurotic organization, and lower ranges, toward the border with the psychoses, describing more severe disturbance (McWilliams & Shedler, 2026, p. 658). The distinction matters for treatment: experienced clinicians suggest more exploratory, insight-oriented work at the higher levels and more supportive, capacity-building work at the lower ones (p. 658; McWilliams, 2011, p. 83). The intense, hard-to-contain countertransference this level evokes is described as especially strong at the lower levels, and people roughly corresponding to those lower levels often first come to medical professionals with bodily complaints (McWilliams & Shedler, 2026, p. 660). Patients at lower borderline levels are also more likely to drop out of treatment (Hörz-Sagstetter et al., 2021). McWilliams calls the borders somewhat arbitrary to begin with, and notes that every neurotic-level person has some borderline tendencies, and vice versa (2011, p. 83). The PDC-3 marks the two borders with scores of 6 and 3 (Gordon & Bornstein, 2025), the STIPO distinguishes three borderline levels (Hörz-Sagstetter et al., 2021), and the M axis spreads the range across M4 to M6 (Lingiardi et al., 2026, pp. 738–740). In the examples above, Kai, rated between the borderline and neurotic levels, is near the higher border, while Donna, as McWilliams first knew her, functioned at the border of the psychoses (McWilliams, 2004, p. 222).
Borderline vs. neurotic
Neurotic-level people can usually recognize a transference as a transference and take an interpretation as information (McWilliams & Shedler, 2026, p. 662). The PDM-3 suggests the decisive difference may be the relative absence of mature defenses under stress (p. 658).
Borderline vs. psychotic
The key difference is reality testing. McWilliams adds two clinical signs: when a therapist gently names a primitive defense, a person at the borderline level shows at least temporary relief, while a person at the psychotic level becomes more agitated; and questions about identity tend to provoke hostility at the borderline level but fear at the psychotic level, because the very sense of existing is at stake (2011, pp. 63–65).
Borderline vs. reactions to stress or trauma
Under enough strain or trauma, anyone can look borderline for a while (McWilliams & Shedler, 2026, p. 654). The level describes an enduring organization, though the PDM-3 also notes that a history of complex trauma can become structured into personality (p. 655).
“A borderline”
The word is often used as an insult or a warning among professionals and the public alike. McWilliams notes the bad reputation borderline clients have among many mental health workers, even though they are not always unpleasant people and usually respond to good treatment (2011, p. 89). This site never uses the word as a noun for a person.

History

As early as 1890, the psychiatrist Rosse wrote of patients who seemed to live in a “borderland” between sanity and insanity (McWilliams, 2011, p. 52). In 1938 Adolph Stern described a “border line group of neuroses” whose members got worse rather than better in standard psychoanalytic treatment (Stern, 1938; McWilliams, 2011, p. 52). Helene Deutsch (1942) described the “as-if” personality, Hoch and Polatin (1949) proposed “pseudoneurotic schizophrenia,” Melitta Schmideberg (1947) described the “stable instability” of such patients, Robert Knight (1953) wrote about “borderline states,” and T. F. Main (1957) described similar patients in hospitals, where staff often divided into opposing camps over them (McWilliams, 2011, pp. 34, 52, 67; McWilliams & Shedler, 2026, p. 655).

Research followed. Grinker, Werble, and Drye (1968) documented a borderline syndrome with a range of severity from the neurotic border to the psychotic border, and Gunderson and Singer (1975) continued the empirical work (McWilliams, 2011, p. 52). Kernberg’s 1967 paper and 1975 book defined borderline personality organization structurally, and Masterson (1976) and others developed developmental accounts (Kernberg, 1967, 1975; McWilliams, 2011, pp. 52, 66).

In 1980, DSM-III made borderline personality disorder an official diagnosis. McWilliams writes that this legitimized a valuable psychoanalytic concept at the price of its meaning as a level of functioning, and that she is “probably fighting a losing battle” in trying to preserve that meaning (2011, p. 53). Kernberg began to distinguish borderline personality organization from borderline personality disorder (p. 53). By the 1990s, attention turned to trauma and attachment, and contemporary research supports constitutional, attachment, and trauma-related contributions (p. 54). Manualized treatments followed, including dialectical behavior therapy, transference-focused psychotherapy, and mentalization-based treatment (p. 71).

The Psychodynamic Diagnostic Manual has wrestled with the double meaning ever since its first edition in 2006. PDM-2 adopted Kernberg’s solution, describing both a borderline level and a borderline personality style, but readers, especially newer clinicians, found this confusing. PDM-3 keeps “borderline” only for the level and calls the style emotionally dysregulated personality (McWilliams & Shedler, 2026, pp. 656, 696).

Sources