GeistHaus

Personality

Frameworks and Thinkers

There is no single theory of personality. Over more than a century, clinicians and researchers have come at the same human puzzle from very different directions: through bodily drives, through defenses, through early relationships, through the sense of self, through attachment, through measurable traits, and through statistics. Each angle lights up something the others miss.

Nancy McWilliams, whose synthesis underlies much of this section, describes herself as more of a synthesizer than a critic, and observes that different clients seem to make different theories relevant: one person brings Kernberg to mind, another sounds like someone Horney described (McWilliams, 2011, p. 22). She also points to work by Robert Stolorow and George Atwood showing how the central themes in a theorist’s own life tend to become the focus of that person’s theory (p. 22). The theories below are best read the same way: as partial maps, each drawn from a particular vantage point.

The page moves roughly in chronological order and ends with the disagreements that are still alive.

Before Psychoanalysis

The four temperaments

The oldest personality typology in the Western tradition comes from ancient medicine. Galen, the Greek physician of the 2nd century AD, developed the notion of temperament from an earlier theory of four bodily fluids, or humors: blood, phlegm, yellow bile, and black bile. The predominance of one humor was said to make a person sanguine, phlegmatic, choleric, or melancholic (Encyclopaedia Britannica, n.d.-c), and the scheme was still current in the European Middle Ages and later (Encyclopaedia Britannica, n.d.-a).

The humors are gone, but the intuition that some differences are inborn has survived. McWilliams opens each of her personality chapters with drive, affect, and temperament, noting that contemporary genetic findings and decades of temperament research make Freud’s attention to innate differences look prescient (McWilliams, 2011, pp. 152–153).

Moral insanity and Kraepelin’s categories

In 1835 James Cowles Prichard used the term “moral insanity” for people we would now call antisocial, who were nonetheless considered to be in touch with reality; the label later gave way to “psychopathy,” “sociopathy,” and “antisocial personality disorder” (McWilliams, 2011, pp. 2, 45). Emil Kraepelin (1856–1926), usually called the father of modern psychiatric classification, grouped mental illnesses into syndromes by careful observation. Freud largely worked within Kraepelin’s categories, and the basic split between neurosis (reality intact) and psychosis (reality lost) shaped early psychoanalytic thinking (pp. 46–47).

Freud and the First Character Types (1908–1933)

Freud’s character papers

Freud’s early theory explained personality through development of the drives. A child passes through oral, anal, and phallic-oedipal phases, and too much frustration or too much gratification at a phase was thought to leave the person “fixated” on its issues. In those early days analysts spoke of oral, anal, or phallic characters (McWilliams, 2011, pp. 23–24).

Freud wrote three short papers that are the starting point for psychoanalytic character study. In “Character and Anal Erotism” (1908) he linked three traits, orderliness, parsimony, and obstinacy, to the anal phase (Abraham, 1927, p. 371; McWilliams, 2011, p. 290). The paper, as Karl Abraham later remarked, “occupied only three pages of a journal” (Abraham, 1927, p. 370), but it launched the idea of what we would now call the obsessive–compulsive character. “Some Character-Types Met With in Psycho-Analytic Work” (1916) described three types: the “exceptions,” who trace their claim to special privileges to an injury in earliest childhood that they felt was unjust; those “wrecked by success,” who fall ill just when a long-cherished wish comes true, because forces of conscience forbid them to enjoy it; and “criminals from a sense of guilt,” whose guilt was present before the misdeed and was relieved by it (Freud, 1916/1925, pp. 319–321, 323–326, 342). In “Libidinal Types” (1931) he proposed erotic, obsessional, and narcissistic types and their mixtures, and suggested that all of them could exist without pathology (“Libidinal types,” n.d.).

Freud’s later structural model, introduced in The Ego and the Id (1923), shifted attention from the contents of the unconscious to the processes that keep them out of awareness, and so set the stage for ego psychology (McWilliams, 2011, p. 27).

Karl Abraham

Karl Abraham expanded the drive-based approach in three papers on character formation: on the anal character (1921), on the influence of oral erotism on character (1924), and on character formation at the genital level (1925), later collected in his Selected Papers (Abraham, 1927). He defined character loosely, as “the sum of his instinctive reactions towards his social environment,” and he insisted that “character is a changeable thing” (pp. 407–408).

Wilhelm Reich and character armor

In Character Analysis (1933), Wilhelm Reich brought the drive approach to personality to its high point. McWilliams notes that his language sounds archaic but that the book is full of insights about character types that still strike a chord (McWilliams, 2011, p. 24). Reich made two lasting contributions. First, he distinguished “symptom neuroses” from “character neuroses”: a discrete problem in an otherwise flexible person versus a personality permeated by neurotic patterns, a distinction that survives in the DSM’s separation of clinical disorders from personality disorders (p. 47). Second, his image of “character armor” captured the idea that rigidity itself is the problem and that health has to do with emotional flexibility (p. 29). His portraits include the self-assured, arrogant “phallic–narcissistic character” and the “masochistic character” (pp. 178, 268), ancestors of today’s narcissistic and masochistic styles.

The drive-based model of character did not last. McWilliams writes that the effort to explain character entirely through fixation “proved disappointing,” though the field kept its developmental sensibility (McWilliams, 2011, p. 24).

Carl Jung’s psychological types

Carl Jung offered a very different typology in Psychological Types (German 1921; English 1923). He described two attitudes, introversion and extraversion, and four basic functions: thinking, feeling, sensation, and intuition. If one function habitually prevails, he wrote, “a corresponding type results,” and each type can be introverted or extraverted (Jung, 1923, p. 14). McWilliams counts Jung’s ideas (in her case, his archetypes) among the less clinically influential psychodynamic formulations that have nonetheless enriched her understanding of personality (McWilliams, 2011, p. 21).

Ego Psychology and the Interpersonal School (1936–1950)

Anna Freud and Heinz Hartmann

With the structural model, character also came to be seen as a characteristic style of defense. To the drive-era question, “At what early maturational phase is this person psychologically stuck?”, ego psychology added another: “What are this person’s typical ways of avoiding anxiety?” McWilliams stresses that the second view did not contradict the first; it offered a new set of ideas and metaphors (McWilliams, 2011, p. 21). Anna Freud’s The Ego and the Mechanisms of Defense (1936) gave that approach its most elegant statement (p. 30). Heinz Hartmann’s Ich-Psychologie und Anpassungsproblem (1939; in English as Ego Psychology and the Problem of Adaptation, 1958) turned attention to adaptation and to a “conflict-free ego sphere,” capacities that develop outside of inner conflict (Hartmann, n.d.).

Ego psychology left the field several durable ideas: the “observing ego” that can step back and comment on experience; the difference between “ego alien” problems, which a person finds strange, and “ego syntonic” ones, which feel like simply the way things are; “ego strength,” the capacity to face unpleasant reality without resorting to denial; and a hierarchy running from immature to mature defenses (McWilliams, 2011, pp. 28–29). It also recognized that health involves being able to use a variety of defenses rather than one (p. 29).

Otto Fenichel

Otto Fenichel’s The Psychoanalytic Theory of Neurosis (1945) covers the defenses, in McWilliams’ phrase, with Fenichel’s “usual thoroughness” (McWilliams, 2011, p. 150), and she calls its chapter on character disorders the standard classical account of the difference between a neurotic symptom and a neurotic character (McWilliams, 2011, p. 68).

Karen Horney

By the middle of the 20th century, a group of American “interpersonal” analysts, including Harry Stack Sullivan, Erich Fromm, Karen Horney, and Frieda Fromm-Reichmann, was developing ideas strikingly parallel to the British object relations school. Compared with their European object-relations counterparts they placed less stress on internalized early figures, and both groups emphasized emotional safety over the delivery of insight (McWilliams, 2011, p. 32). Fromm-Reichmann’s famous summary: “The patient needs an experience, not an explanation” (p. 32).

Horney departed from some of Freud’s basic principles and proposed an environmental and social basis for personality and its disorders (Encyclopaedia Britannica, n.d.-b). She traced neurotic patterns to “basic anxiety,” the insecurity of feeling unsafe, unloved, and unvalued. In Our Inner Conflicts (1945) and later Neurosis and Human Growth (1950) she described three strategies for managing it: moving toward, against, or away from other people, with corresponding self-effacing, expansive, and resigned solutions. She also contrasted the “real self” with an idealized self-image that people compulsively pursue (Paris, n.d.). The three movements map loosely onto the dependent, narcissistic or paranoid, and schizoid styles of later systems; that mapping is ours, not Horney’s.

Erich Fromm’s character orientations

In Man for Himself (1947), Fromm described five character orientations: four nonproductive ones (receptive, exploitative, hoarding, and marketing) and a productive orientation he saw as the healthy ideal. For Fromm, character was shaped by individual choice but also by wider social and economic forces (Braune, 2017). McWilliams notes that Fromm was among the psychoanalytic social critics who argued that modern life reinforces narcissistic concerns (McWilliams, 2011, p. 177).

Erik Erikson

Erikson’s Childhood and Society (1950) recast Freud’s stages in terms of interpersonal tasks: the oral phase became a question of basic trust versus mistrust, the anal phase autonomy versus shame and doubt, and the oedipal phase initiative versus guilt. He extended development across the whole life span and across cultures (McWilliams, 2011, pp. 24–25). His concept of an integrated identity (Erikson, 1968) became one of the central markers of personality health (p. 57), and the PDM-3, drawing on Kernberg and on Erikson, still distinguishes a normal adolescent identity crisis from the identity diffusion that signals personality difficulty (Tanzilli & Sharp, 2026, p. 469).

Object Relations (1930s–1970s)

Melanie Klein

In Britain, analysts working with children and with more disturbed adults found they needed a new language. Melanie Klein described two basic states of mind. In the “paranoid–schizoid position,” other people are not yet fully appreciated as separate. In the “depressive position,” the person understands that others have separate minds outside one’s control (McWilliams, 2011, pp. 25, 31). Later writers stressed that we all move back and forth between these states from moment to moment (p. 25). Klein’s 1946 paper “Notes on Some Schizoid Mechanisms” was the first to describe projective identification (p. 113), and the PDM-3 still cites it in explaining splitting of the self and of others into all good and all bad (McWilliams & Shedler, 2026, pp. 658–659).

Fairbairn and Guntrip

W. R. D. Fairbairn rejected Freud’s biology outright: people, he argued, seek relationships more than drive satisfaction. A baby wants not just milk but the warm experience of being nursed (McWilliams, 2011, p. 31). Object relations theory, “object” being the analysts’ unlovely word for a person, asked what the important people in a child’s world had been like and how their images live on inside the adult (pp. 31–32). Fairbairn and Harry Guntrip wrote with special insight about schizoid states, which Fairbairn described as “love made hungry” (p. 199). Guntrip, in Schizoid Phenomena, Object Relations and the Self (1969), named the schizoid person’s “in and out programme”: unable to be either inside a relationship or outside it without feeling at risk (p. 201).

Donald Winnicott

Winnicott’s provocative remark that there is no such thing as a baby, only a baby with someone caring for it, anticipated the later relational turn (McWilliams, 2011, p. 39). He also wrote on “Ego Distortion in Terms of True and False Self” (1960). His ideas of the holding environment and of mirroring helped redefine therapy for people whose sense of self is fragile (p. 177). His frank 1949 paper “Hate in the Countertransference” was part of a broader shift, one of object relations theory’s most important contributions, that turned the therapist’s own feelings from an obstacle into a source of understanding (pp. 34–36). Heinrich Racker, a South American analyst influenced by Klein, added a useful distinction: in “concordant” countertransference the therapist feels what the patient once felt toward an early figure, and in “complementary” countertransference the therapist feels what that early figure felt toward the patient as a child (p. 34).

Margaret Mahler

Mahler and her colleagues, in The Psychological Birth of the Human Infant (1975), traced how a toddler moves from a symbiotic closeness with the caregiver through subphases of “hatching,” “practicing,” and “rapprochement” toward a separate self, a process that reaches its first resolution around age 3 (McWilliams, 2011, p. 25). James Masterson later linked the stormy back-and-forth of the borderline level to the rapprochement subphase, when a 2-year-old alternates between “I can do it myself!” and collapsing in tears (pp. 66–67). McWilliams notes that whether or not his ideas about cause are right, his observations make sense of the dilemmas he described (p. 67).

Styles and Levels (1930s–1980s)

The borderline concept

From the late 1930s, clinicians described patients who seemed too troubled to be called neurotic yet too anchored in reality to be called psychotic. In McWilliams’ words, they were “too sane to be considered crazy, and too crazy to be considered sane” (McWilliams, 2011, p. 52). A series of writers named and studied this middle ground, among them Adolph Stern, Helene Deutsch with her “as-if personality,” Robert Knight, Roy Grinker’s research group, and John Gunderson and Margaret Singer (McWilliams, 2011, p. 52; McWilliams & Shedler, 2026, p. 655). The result was a spectrum of severity running from healthy or “normal” through neurotic and borderline to psychotic (McWilliams, 2011, p. 43).

David Shapiro’s Neurotic Styles (1965)

Shapiro’s short book examined four styles, obsessive–compulsive, paranoid, hysterical, and impulsive, in terms of their characteristic ways of thinking and perceiving, forms of emotional experience, and modes of activity (Shapiro, 1965). In effect, Shapiro treated a style as a whole way of thinking, perceiving, feeling, and acting, not just a set of defenses. The PDM-3 names it, with Kernberg’s levels and McWilliams’ work, as one of the formulations that shaped its own system (Lingiardi & McWilliams, 2026, p. 1), and it still cites Shapiro for the “relative rigidity” of neurotic-level people (McWilliams & Shedler, 2026, p. 657).

Otto Kernberg’s structural model

Otto Kernberg’s 1967 paper “Borderline Personality Organization” and his books Borderline Conditions and Pathological Narcissism (1975) and Severe Personality Disorders (1984) turned the borderline idea into a systematic model. He defined borderline personality organization by three features: identity diffusion, a predominance of primitive defenses, and fragile but intact reality testing (McWilliams & Shedler, 2026, p. 655). He proposed “adequacy of reality testing” as the way to tell borderline from psychotic organization (McWilliams, 2011, p. 65), and his structural interview, later developed by his colleagues into the Structured Interview for Personality Organization (STIPO), made the levels assessable (p. 55). Kernberg also described “malignant narcissism,” narcissism suffused with sadistic aggression, which he placed on a continuum with frankly psychopathic personality (McWilliams & Shedler, 2026, p. 688). Transference-focused psychotherapy, one of the evidence-based treatments for severe personality problems, is associated with his group (p. 661).

When the DSM adopted “borderline personality disorder” in 1980, it drew heavily on Gunderson’s research with a group most analysts would have seen as having a histrionic style at the borderline level. Kernberg then had to distinguish borderline personality organization (a level) from borderline personality disorder (a type) (McWilliams, 2011, pp. 52–53). The PDM-3 resolves the confusion by calling the DSM pattern the emotionally dysregulated style (McWilliams & Shedler, 2026, p. 656).

Heinz Kohut’s self psychology

By the 1960s many therapists were seeing patients who did not feel full of harsh inner critics but empty: unsure of their values, dependent on outside reassurance, and often outwardly composed. Heinz Kohut built a new theory around them in The Analysis of the Self (1971), The Restoration of the Self (1977), and How Does Analysis Cure? (1984). He emphasized the normal need to idealize others and the damage done when a child grows up without people who can first be idealized and then gently de-idealized (McWilliams, 2011, pp. 36–37). Self psychologists coined the term “selfobjects” for the people who support our self-esteem through affirmation and admiration; everyone needs them, and a person with narcissistic difficulties needs them so much that other aspects of relationship fade (pp. 182–183). They also described several kinds of “selfobject transference” in therapy, including mirroring and twinship patterns (p. 188).

Kohut’s influence reached well beyond narcissism. To the traditional questions about defense, “Of what is this person afraid? When afraid, what does this person do?”, self psychology added two more: “How vulnerable is this person’s self-esteem? When it is threatened, what does he or she do?” (McWilliams, 2011, p. 38). He also drew attention to the importance of the therapist acknowledging empathic failures (p. 190).

The Kohut–Kernberg debate

Through the 1970s and 1980s, Kohut and Kernberg offered competing accounts of narcissism. Kohut saw it developmentally: a person’s growth went along normally and then ran into trouble with ordinary needs to idealize and de-idealize. Kernberg saw it structurally: something went awry very early, leaving a pathological self-structure and entrenched primitive defenses that differ in kind from normal self-regard (McWilliams, 2011, pp. 188–189). McWilliams offers an image: Kohut’s narcissistic person is a plant stunted by too little water and sun, Kernberg’s a plant that has mutated into a hybrid (p. 189).

The treatments followed the theories. Kohut’s followers recommend accepting idealization or devaluation and offering unwavering empathy; Kernberg recommends tactful but insistent confrontation of grandiosity and interpretation of defenses against envy and greed (McWilliams, 2011, p. 189). Kernberg has suggested that Kohut’s approach may suit more disturbed patients as a form of supportive therapy, even though Kohut worked mostly with high-functioning people. McWilliams’ verdict is pragmatic: most analysts she knows have some patients for whom Kohut fits and others for whom Kernberg fits, and “the jury is still out” (p. 189).

Attachment, Development, and Relationship (1969–2000s)

Attachment research

Attachment theory, which McWilliams calls “the conceptual baby of the psychoanalyst John Bowlby” (McWilliams, 2011, p. xii) and which Bowlby set out in Attachment (1969), gave the field a way to study early relationships directly. Bowlby’s idea of “inner working models” of relationships is one of the many concepts that resemble psychoanalytic notions of personality as recurring internal scripts (Lingiardi & McWilliams, 2026, p. 7). Mary Ainsworth and colleagues’ Strange Situation procedure (1978) classified infants’ patterns of attachment, including avoidant and ambivalent/resistant patterns (Speranza & Steele, 2026, pp. 122–123), and Mary Main and Judith Solomon (1986) identified a disorganized/disoriented pattern, sometimes called “type D,” in which the attachment figure is at once a source of safety and of fear (McWilliams & Shedler, 2026, p. 696). People at more severe levels of personality organization typically show one of the insecure attachment styles when tested (p. 656).

Attachment research also connected generations. In a well-known study, parents’ capacity for reflective functioning, assessed in Adult Attachment Interviews before their first child was born, predicted the child’s attachment pattern a year or more later (Speranza & Steele, 2026, p. 117).

Sidney Blatt’s two configurations

Sidney Blatt distinguished two kinds of depression. In the introjective kind, the person feels bad, guilty, and self-critical: “I’m not good enough, I’m flawed.” In the anaclitic kind, from a Greek word meaning “to lean on,” the person feels empty, lonely, and in need of connection (McWilliams, 2011, pp. 38, 237). Blatt later generalized the two poles beyond depression as “self-definition” and “self-in-relationship.” Everyone has both needs, and a healthy balance between them is part of mental health (p. 237; Blatt, 2008). The distinction has practical consequences: an overtly sympathetic stance can relieve a person in anaclitic pain but make a self-critical, introjective person feel misunderstood (McWilliams, 2011, p. 38). The PDM-3 builds Blatt’s model into its description of the depressive style (McWilliams & Shedler, 2026, p. 668).

Relational psychoanalysis

Jay Greenberg and Stephen Mitchell’s Object Relations in Psychoanalytic Theory (1983) contrasted drive and ego models with relational ones and set off what came to be called the “relational turn” (McWilliams, 2011, p. 39). Relational analysts doubt that therapists can be neutral observers and see therapy as a mutual, co-constructed relationship in which both people’s unconscious lives play a part. They have been more interested in process than in hypothesized structures such as character; many worry that describing personality as fixed ignores how much our experience is constructed in the moment and how much it depends on our current “self-state” (p. 39). Their work on multiple self-states and dissociation has reshaped thinking about trauma (p. 40).

Other theories within the broad psychoanalytic family have also shaped ideas about character, among them the work of Alfred Adler and Otto Rank, Henry Murray’s “personology,” Silvan Tomkins’ “script theory,” control–mastery theory, evolutionary models, contemporary gender theory, and the work of Jacques Lacan (McWilliams, 2011, p. 40). Not everyone builds them in: the PDM-3 deliberately leaves out Lacanian diagnosis, partly because its categories do not mesh easily with developmental and dimensional approaches and partly because empirical support is still limited (Lingiardi & McWilliams, 2026, p. 2).

Cyclical psychodynamics

Paul Wachtel, writing from the 1970s on, offered a different answer to the question of why patterns persist. Rather than treating a character trait as a residue of early life sealed off from later experience, he described it as a loop kept going in the present: an anxiety leads to a way of behaving, the behavior draws responses from other people that confirm the anxiety, and the loop runs again (Wachtel, 1997). The approach is psychodynamic in its attention to conflict and defense and integrative in its borrowing from behavioral and family-systems work. Its practical companion, Therapeutic Communication (Wachtel, 2011), is about wording: how the same observation can be phrased as permission or as rebuke, and why the difference decides whether it can be used. See vicious circles and accomplices.

Mentalization

Peter Fonagy and colleagues described how children move from a mode in which inner and outer reality are equated, through a “pretend mode,” toward the capacity to mentalize: to understand behavior, one’s own and others’, in terms of thoughts, feelings, and intentions (McWilliams, 2011, p. 26; Fonagy et al., 2002). Fonagy linked difficulties in this “reflective function” to insecure attachment and to borderline psychology (McWilliams, 2011, p. 64). With Anthony Bateman he developed mentalization-based treatment (Bateman & Fonagy, 2004), one of several evidence-based therapies for severe personality problems (McWilliams & Shedler, 2026, p. 661). The PDM-3 lists mentalization as one of its thirteen mental capacities (Lingiardi et al., 2026, p. 711).

The Psychiatric Manuals and the Empirical Turn (1952–2012)

DSM-I and DSM-II

The first Diagnostic and Statistical Manual (DSM-I, 1952) had no category for hysterical personality, folding some of its features into an “emotionally unstable personality” (Novais et al., 2015). DSM-II (1968) was strongly shaped by psychoanalysis (Kapadia et al., 2020; Novais et al., 2015). McWilliams writes that DSM-II accepted both depressive and cyclothymic personality disorders (McWilliams, 2011, p. 256); depressive personality disorder as a formally defined diagnosis, however, first appeared only in DSM-IV (1994), in an appendix of categories needing further study (Ørstavik et al., 2012).

DSM-III and Axis II (1980)

DSM-III (1980) took a deliberately atheoretical, descriptive approach and introduced a multiaxial system with explicit diagnostic criteria (Kapadia et al., 2020). Personality disorders were placed on their own axis, Axis II, described as discrete types, and grouped into three clusters (Millon, 1981; Novais et al., 2015). “Hysterical” became “histrionic,” partly because the old descriptions of the “hysterical female” were seen as pejorative (Novais et al., 2015), and borderline personality disorder entered the manual (McWilliams, 2011, pp. 52–53).

The PDM-3 argues that to include personality styles in a manual of “disorders,” DSM-III’s framers described them in their most extreme forms, and that many clinicians consequently learned to ignore personality unless it met criteria for a disorder (McWilliams & Shedler, 2026, p. 653). Later revisions kept the categorical structure: DSM-III-R (1987) and DSM-IV (1994) continued the clustered types (Novais et al., 2015). Sadistic personality disorder appeared as a provisional diagnosis in DSM-III-R and disappeared in DSM-IV (McWilliams & Shedler, 2026, p. 694), A self-defeating (masochistic) personality disorder, one of the major controversies of the DSM-III-R revision (Kass et al., 1989), was likewise proposed in DSM-III-R and later removed (Huprich et al., 2006); the American Psychiatric Association considered but did not adopt it for DSM-IV (McWilliams, 2011, p. 268). Depressive personality remains absent from the DSM today (McWilliams & Shedler, 2026, p. 667), and McWilliams counts the disappearance of the “time-honored” depressive personality diagnosis among the arbitrary consequences of the post-1980 system (p. 10).

Theodore Millon

Theodore Millon was a central figure in the Axis II era. His book Disorders of Personality: DSM-III, Axis II (1981) accompanied the new system, and an obituary credits him with separating personality disorders from clinical disorders such as depression and anxiety. He was also known as one of the few people who tried to bridge the different schools of thought (Watts, 2014). Millon’s descriptions of an “aggrieved” self-defeating pattern and of sadistic personality are still cited by psychodynamic writers (McWilliams, 2011, p. 268; McWilliams & Shedler, 2026, p. 694).

Lorna Smith Benjamin and SASB

Lorna Smith Benjamin’s Structural Analysis of Social Behavior (Benjamin, 1974) took an interpersonal route. It is a circumplex model that describes interactions along dimensions of affiliation (love versus attack) and interdependence (emancipation versus control), with three different “surfaces”: focus on the other, focus on the self, and the “introject,” the way a person treats themselves as they were habitually treated by others (Margola et al., 2020). It gave researchers a precise language for the relational patterns that psychodynamic clinicians described more intuitively.

Trait psychology and the five-factor model

Academic personality psychology developed largely in parallel, measuring traits rather than inferring themes. Its best-known product is the five-factor model, whose broad traits of extraversion, neuroticism, openness to experience, agreeableness, and conscientiousness were derived by factor analysis (McCrae & John, 1992; Lingiardi & McWilliams, 2026, p. 7). Trait research has heavily influenced recent official systems (Lingiardi & McWilliams, 2026, p. 7). Psychodynamic writers respond that a trait list can miss the organizing theme: McWilliams stresses that analytic theories describe “themes and dynamisms, not traits,” and, quoting Philip Slater, that traits and their opposites coexist in the same person when they are of any intensity (McWilliams, 2011, pp. 40–41).

Oldham and Morris’s normal styles (1990)

Psychiatrist John Oldham and journalist Lois Morris took the DSM categories in the opposite direction. In The Personality Self-Portrait (1990; revised as The New Personality Self-Portrait, 1995) they described normal personality styles; the online test that grew out of the book now assesses fourteen of them (NPSP25, n.d.). Oldham explains that they “conceptually turned down the volume” of each official or proposed disorder until it described an adaptive style, and he argues that a personality disorder is not an on–off matter but an extreme version of a style that has turned from asset into liability (Oldham, n.d.).

Westen, Shedler, and the SWAP

Drew Westen and Jonathan Shedler tried to bridge clinical wisdom and statistics. Their Shedler–Westen Assessment Procedure (SWAP-200) lets a clinician describe a patient in rich clinical detail but in a systematic, quantifiable form, based on the clinician’s observation rather than on the patient’s answers to direct questions (Westen & Shedler, 1999a). When hundreds of experienced clinicians used it to describe patients with personality disorders, the patients sorted into naturally occurring groups that only partly matched Axis II, including a large “dysphoric” group the DSM did not recognize (Westen & Shedler, 1999b). A later national sample of 1,201 patients, each described by their clinician, produced ten personality syndromes in three higher-order clusters, internalizing, externalizing, and borderline-dysregulated, and a paragraph-length prototype for each (Westen et al., 2012). Those prototypes are the backbone of the PDM-3 personality styles (McWilliams & Shedler, 2026, p. 666).

Dimensional Systems (2006–2026)

The Psychodynamic Diagnostic Manual (2006, 2017, 2026)

Prompted by the limits of descriptive, categorical classification, which says little about how patients actually experience their suffering, Stanley Greenspan spearheaded the first Psychodynamic Diagnostic Manual (PDM Task Force, 2006; Mundo et al., 2026, p. 761). A second edition followed in 2017 and a third in 2026, edited by Vittorio Lingiardi and Nancy McWilliams (Lingiardi & McWilliams, 2017, 2026). The PDM calls itself a “taxonomy of people” rather than a “taxonomy of disorders” (Lingiardi & McWilliams, 2026, p. 3).

The PDM-3 assesses adults on three axes: the P Axis of personality level and style, the M Axis of mental capacities, and the S Axis of how symptoms are subjectively experienced (Lingiardi & McWilliams, 2026, p. 10). Its personality diagnoses are dimensional and holistic: a clinician rates how closely a person matches each prototype rather than counting criteria (p. 5). The third edition reserves “borderline” for a level of organization, renames the DSM-style borderline pattern emotionally dysregulated, and expands the M Axis from 12 to 13 capacities (p. 12). It also covers personality across the life span, with separate chapters on adolescents and older adults (pp. 9–10). The first PDM had classified dissociative identity disorder as a “dissociative personality” (see the dissociative style); later editions returned it to the dissociative disorders (Mundo et al., 2026, p. 831).

DSM-5 and its alternative model (2013, 2022)

DSM-5 (2013) dropped the multiaxial system (Kapadia et al., 2020) but kept the ten categorical personality disorders in its main section. In Section III it printed an Alternative Model for Personality Disorders (AMPD), which asks two questions: how impaired is the person’s functioning of self (identity, self-direction) and relationships (empathy, intimacy), and which of 25 pathological trait facets, grouped into five domains (negative affectivity, detachment, antagonism, disinhibition, and psychoticism), are present? A diagnosis requires at least moderate impairment plus pathological traits. The model defines six specific disorders and a “trait specified” category for everyone else (Clark, 2025). DSM-5-TR (2022) kept this arrangement (Lingiardi & McWilliams, 2026, p. 5).

The PDM-3 calls the AMPD “a significant advance” because it is dimensional, while noting that trait models, even dimensional ones, may miss thematic differences between people (Lingiardi & McWilliams, 2026, p. 7). In our reading, its first criterion, level of personality functioning, is a close cousin of the psychodynamic idea of level of organization. As of 2025, a revision process was under way that seemed likely, but far from guaranteed, to move a dimensional model into the DSM’s main section (Clark, 2025).

HiTOP (2017)

The Hierarchical Taxonomy of Psychopathology (HiTOP) is a large collaborative attempt to build a dimensional classification by applying statistical methods to large data sets (Lingiardi & McWilliams, 2026, p. 8). It groups symptoms that tend to occur together into components and syndromes, combines co-occurring syndromes into broad spectra, and treats everything dimensionally, addressing the arbitrary boundaries and heavy overlap of traditional categories (Kotov et al., 2017). As the PDM-3 summarizes it, psychopathology is organized along broad dimensions including internalizing, thought disorder, disinhibited and antagonistic externalizing, and detachment. The PDM-3 sees HiTOP as a potentially significant complement to categorical diagnosis but criticizes it, like the AMPD, for relying on measurable traits rather than inferable themes, and for requiring psychometric tools that are hard to use in ordinary practice (Lingiardi & McWilliams, 2026, p. 8).

Other alternatives to the DSM and ICD

The PDM-3 introduction surveys several other ways of classifying psychological problems. It counts as close cousins of its own approach the SWAP, the Structured Interview of Personality Organization developed by Kernberg’s group, the Operationalized Psychodynamic Diagnosis (OPD) system, and Blatt’s model of anaclitic and introjective configurations, and it notes the growing neuropsychoanalysis movement, which links personality to basic emotional systems in the mammalian brain (Lingiardi & McWilliams, 2026, pp. 6–7). Further afield are the U.S. National Institute of Mental Health’s Research Domain Criteria (RDoC), developed around 2010 to classify problems by impaired brain systems; network approaches, which treat symptoms as interacting nodes rather than signs of hidden disorders; and the British Psychological Society’s Power Threat Meaning Framework, which asks people how power has operated in their lives and how they survived it. The PDM-3 shares the last one’s respect for patients as subjects but is unwilling to discard psychiatric diagnosis altogether (pp. 8–9).

ICD-11 (2022)

The World Health Organization’s ICD-11, adopted in 2019 and in effect since January 2022 (World Health Organization, 2019), made the most radical change. ICD-10 had a categorical list much like the DSM’s (Clark, 2025). ICD-11 abolished the types. A clinician first decides whether there is a personality disorder and rates it mild, moderate, or severe; then, optionally, notes prominent trait domains (negative affectivity, detachment, dissociality, disinhibition, and anankastia) and a borderline pattern. Below the threshold, “personality difficulty” records milder, intermittent problems and is explicitly not a mental disorder (Bach & First, 2018). Severity, in other words, became the core of the diagnosis, much as the PDM-3 makes level of organization the first step of its personality diagnosis (McWilliams & Shedler, 2026, p. 656).

Common Threads and Real Disagreements

Across a century of disagreement, a few convictions run through most of the frameworks above, and the PDM-3 states them plainly: personality matters for understanding symptoms; severity and style are different questions; personality health and disturbance lie on a continuum rather than in sharp boxes; and genes, temperament, early experience, and attachment all contribute (McWilliams & Shedler, 2026, pp. 652–656). The real arguments that remain are worth knowing about.

Developmental stages versus many causes

Early psychoanalytic theories tied each level or type of personality to a particular stage of childhood. Many contemporary writers find stage explanations too reductive and only questionably supported by evidence; developmental maps, in one writer’s words, “should not be taken literally” (McWilliams, 2011, pp. 26–27). Infant research has found more competence in early infancy than stage models assumed, and critics point out that models of “normal development” carry cultural assumptions (p. 44). The PDM-3 now says that viewing level of organization as fixation at a stage “holds little currency,” pointing instead to a confluence of genetics, temperament, early experience, and attachment (McWilliams & Shedler, 2026, p. 656). McWilliams keeps a developmental axis anyway, partly because the idea that any of us could have gotten stuck under unfortunate circumstances invites compassion (McWilliams, 2011, p. 27).

Categories versus dimensions

Is a personality disorder something you have or don’t, or a matter of degree? The PDM-3 argues that virtually all psychological phenomena are inherently dimensional and that DSM-5-TR, despite its alternative model, remains predominantly categorical; ICD-11 is more dimensional (Lingiardi & McWilliams, 2026, p. 5). Advocates of ICD-11 point out that once the ten types are gone, the overlapping “comorbidity” of personality diagnoses largely disappears (Bach & First, 2018). The large research literature built on the categories is one reason they persist; the PDM-3 itself keeps citing research on DSM borderline personality disorder for that reason (McWilliams & Shedler, 2026, p. 656).

Conflict versus deficit

Is a person’s difficulty best understood as an inner conflict, wishes at war with fears or prohibitions, or as a deficit, something that never developed? The PDM-3 describes the problems of people at the neurotic level largely in terms of inner discord or conflict (McWilliams & Shedler, 2026, p. 657). Kohut and others reframed many problems as developmental arrest and deficit (McWilliams, 2011, p. 177). The difference shapes therapy: one writer McWilliams cites suggests that a more “maternal,” holding voice fits deficits while a more interpretive voice fits conflicts, and that most people have both, so therapy moves between them (McWilliams, 2004, p. 146). The PDM-3 makes a similar move for the schizoid style, distinguishing a deficit-based version from a higher-functioning one better understood in terms of inner conflict (McWilliams & Shedler, 2026, p. 681).

Kohut versus Kernberg

The debate over narcissism is a special case of the conflict–deficit question with a practical edge: empathize with the wound or confront the grandiosity? It has not been settled. Many clinicians use both, depending on the person and on how severe their difficulties are (McWilliams, 2011, pp. 188–189).

Prototypes versus checklists

Traditional DSM diagnosis counts criteria: a person qualifies by meeting a set number of listed features. The PDM-3 instead asks how closely a person resembles a paragraph-length prototype, on the grounds that recognizing a personality is like recognizing a face, a matter of the whole pattern rather than a tally of features (McWilliams & Shedler, 2026, p. 666). It prefers “prototypic” to “polythetic” diagnosis because clinicians need to understand their patients’ inner lives (Lingiardi & McWilliams, 2026, p. 7). McWilliams warns that the checklist approach makes some serious misreadings easy, such as mistaking a hysterical person for a narcissistic one (McWilliams, 2011, p. 153).

Styles versus disorders

Freud’s libidinal types, Oldham and Morris’s styles, and the PDM all insist that everyone has a personality and that most personalities are not disorders (“Libidinal types,” n.d.; Oldham, n.d.; McWilliams & Shedler, 2026, p. 652). The official manuals, built to classify illness, start from the other end, though ICD-11’s “personality difficulty” is a step toward naming the space between (Bach & First, 2018). Where exactly to draw the line remains a judgment call: McWilliams calls it “quite blurry” (McWilliams, 2011, p. 51). And some relational analysts question the premise that personality is a stable “thing” at all, preferring to speak of shifting self-states that arise in different relationships (p. 39).

Sources