The DSM’s Reconnection to Psychoanalytic Theory through the Alternative Model for Personality Disorders

Early Conceptualizations of Personality in Psychiatric Nosology

By many accounts, Philippe Pinel’s Traité medico-philosophique sur l’aliénation mentale ou la manie (1801) was the first piece of literature to include personality pathology in psychiatric nosology with his introduction of the category of mania without delusion (Berrios 1996). Mania without delusion—manie sans délire—described individuals with tendencies for agitation and impulsive violence without demonstrating abnormalities in perception, memory, judgment, or understanding (Pinel 1801). A number of physicians and psychologists would follow Pinel in describing a category of pathology characterized by peculiar behavior and the abnormal expression of emotions in the absence of delusion. Among these categories were Jean-Étienne Dominique Esquirol’s monomanie raisonnante (1838) and James Cowels Prichard’s moral insanity (Whitlock 1967), both of which identified a similar condition of personality pathology. Many other models or systems for classifying normal and abnormal personality materialized during the late nineteenth century and continued to emerge through the early twentieth (see, e.g., Gauchet and Lambert 1959; Leontyev 2008; Ribot 1896). Even with numerous conceptualizations of personality, the classification systems of the time were quite simplistic. For instance, the 1840 U.S. census grouped all mental illness under a single compressed category—insane and idiotic (Manderscheid et al. 1986). As psychiatric diagnosis developed, personality pathology was generally overlooked by early classification systems; many classification systems, and psychiatry as a field, approached personality pathology as a single condition, such as the poorly defined concept of degeneracy (Crocq 2013). Nevertheless, at the beginning of the twentieth century, several individuals sought to understand personality and offer a better framework for organizing personality and individual differences. Two prominent models of personality during this time were those introduced by Sigmund Freud and the personality types proposed by Emil Kraepelin and his followers.

Models of Personality Presented by Freud and Kraepelin

Freud’s initial unified framework of mental life, the topographic model, divided the mind into three regions: conscious, preconscious, and unconscious (Freud 1900, 1911). Within this model, personality and its development were believed to be highly influenced by the affect-laden unconscious, much more so than by any contributions made by the conscious or preconscious parts of the mind (Bornstein, Maracic, and Natoli 2018). Freud eventually moved away from his topographical model and shifted toward a structural model of human mental life (Freud 1923). This new structural model postulated that personality is derived from the interplay of id, ego, and superego (Freud 1933, 1940). Neither Freud’s topographical model nor his structural model directly categorized personality into specific configurations or disorders. Yet the theory underlying Freud’s structural model (e.g., that the power and influence of id, ego, and superego differ across individuals) has permitted unique personality styles to be roughly organized in terms of variations in the level of power and control each of these psychic structures maintains (Bornstein, Maracic, and Natoli 2018).

As Freud’s initial model of human mental life was being disseminated, Emil Kraepelin had been studying individuals with impaired social functioning and subsequently introduced to psychiatry a number of personality categories based on his findings (Crocq 2013). However, unlike Freud, who acknowledged the influence that experience and early social environment had on the development of personality (Freud 1933, 1940), Kraepelin minimized the influence of early life experience and postulated that personality is the product of an inborn psychological defect (Crocq 2013). Theorized to involve problematic affect and emotions, yet intact cognitive capabilities, Kraepelin’s psychopathic personalities were initially divided into four types: (1) the born criminal, modeled on James Cowels Prichard’s moral insanity (Whitlock 1967); (2) the irresolute and weak-willed, those unwilling or unable to engage in sustained, long-term work; (3) pathological liars and swindlers, characterized by inconsistent memory, hyperactive imaginations, emotional instability, and weak willpower; and (4) “pseudoquerulants,” who are impaired by their paranoid beliefs (Kraepelin 1904). Eleven years later, Kraepelin (1915) revised and expanded his list of personality types to seven: (1) the excitable; (2) the irresolute; (3) persons following their instincts; (4) eccentrics; (5) pathological liars and swindlers; (6) enemies of society; (7) the quarrelsome. Kraepelin’s personality types largely fail to correspond with the personality disorders enumerated in DSM-5. Yet one can effortlessly perceive the parallel between the contemporary DSM system and Kraepelin’s categorical organization of personality types and his focus on specific symptoms that deviate from personality features found in “normal” people.

The Movement Toward Standardized Systems of Personality Diagnosis and the Current Dsm

Influenced by the work of Freud (1900, 1911, 1933, 1940), Kraepelin (1904, 1915), and others, psychiatric diagnosis and the classification of personality in the U.S. during the first half of the twentieth century had made significant strides since once grouping all mental illness under a single category. Still, the classification of disorders significantly varied across the country; the majority of institutions working with the mentally ill employed their own “in-house” systems for diagnosing patients (Surís, Holliday, and North 2016). Although a number of “validated” diagnostic systems were also in existence during this time (e.g., Standard Classified Nomenclature of Disease, an Armed Forces system, and a system developed by the Veterans Administration; Fischer 2012), none of these systems were consistent with the ICD’s new mental disorders section (World Health Organization 1949). In an effort to better standardize psychiatric diagnosis and the language used to describe mental health, the APA Committee on Nomenclature and Statistics proposed the first edition of the DSM in 1952. This first edition was heavily influenced by psychoanalytic theory (Shorter 2005) and went a long way toward meeting the need for improved diagnostic nomenclature. Yet the diagnostic categories included in this first edition were poorly defined and did not strongly correspond with the commonly used ICD. Discussions to revise DSM-I through an international collaborative effort were held, and a new DSM system based on sound psychoanalytic theory (DSM-II, 1968) was designed to correspond with the ICD-8 (World Health Organization 1965). Unfortunately, diagnostic reliability and validity were never established for these early editions, and there was poor differentiation between a number of the included personality disorders and the corresponding neuroses (e.g., hysterical personality and neurosis). The combination of these issues and various developments in the field of psychiatry prompted a profound paradigm shift in the diagnosis and categorization of personality pathology, a shift that involved the DSM’s unfortunate removal of psychoanalytic theory.

DSM-III (1980), in a radical departure from previous editions, defined personality pathology in terms of a categorical, hierarchical taxonomic system closely aligned with Kraepelinian conceptualizations of mental illness. A second noteworthy change introduced in DSM-III was the presentation of specific diagnostic criteria for each personality disorder and a greater neutrality with respect to etiological theories. These changes were intended to make the DSM more useful to researchers and secondary mental health providers (insurance providers among them) by presenting personality disorders in terms of distinct, noninferential categories based on the presence versus absence of specific symptoms. Both of these changes, influenced by neo-Kraepelinian psychiatrists, seemingly diminished the DSM’s clinical utility, particularly with regard to its ability to inform comprehensive case formulation and treatment planning (Vanheule 2017).

DSM-III’s categorical system, which was retained in DSM-IV and DSM-5, defines ten personality disorders in terms of seven to nine symptoms, depending on the specific disorder, that cover areas of cognition, affectivity, interpersonal functioning, and impulse control. According to this system, a personality disorder diagnosis is to be assigned when an individual maintains a set number of symptoms that result in clinically significant impairment. Regrettably, these polythetic diagnostic thresholds were arbitrarily established (Widiger 2001), and the large assortment of symptom combinations within each personality disorder allowing for diagnosis (see Table 1) suggests a remarkable potential for problematic heterogeneity. Such heterogeneity among persons given identical diagnoses likely contributes to significant, long-standing complications in clinical practice (e.g., widely recognized limitations to diagnostic validity, disproportionate comorbidity, diagnostic unreliability, and poor convergent and divergent validity of criteria sets). Even with these gross limitations, relatively minimal changes were made over the next few decades to DSM’s conceptualization of personality or to its approach to diagnosing personality pathology. In fact, as noted, DSM-III’s reconceptualization of personality disorder diagnosis is still largely utilized in DSM-5.

OPEN IN VIEWER

Table 1.

Symptom configurations for DSM-5 Categorical Personality Disorder Diagnosis

Personality Disorder	Diagnostic Criteria	Symptoms	Possible combinations with minimum number of symptoms required	Total possible combinations	Diagnosis possible with no shared symptoms?
Paranoid	4 (or more)	7	35	64	No (share 1)
Schizoid	4 (or more)	7	35	64	No (share 1)
Schizotypal	5 (or more)	9	126	256	No (share 1)
Antisocial	3 (or more)	7	35	99	Yes
Borderline	5 (or more)	9	126	256	No (share 1)
Histrionic	5 (or more)	8	56	93	No (share 2)
Narcissistic	5 (or more)	9	126	256	No (share 1)
Avoidant	4 (or more)	7	35	64	No (share 1)
Dependent	5 (or more)	8	56	93	No (share 2)
Obsessive-compulsive	4 (or more)	8	70	163	Yes

After DSM-III, the APA’s next major modification or, perhaps more appropriately, reformulation of personality classification was the inclusion of the AMPD in DSM-5. Including those mentioned above, the DSM’s long-standing categorical, hierarchical taxonomic system for conceptualizing personality pathology (and the disregard for nonpathological aspects of personality) had drawn a vast amount of criticism over the decades (see Clark, Livesley, and Morey 1997; Cloninger 2000; Livesley 2003; Oldham and Skodol 2000; Rounsaville et al. 2002; Skodol 2014; Tyrer 2001; Vanheule 2017; Widiger 1993; Widiger and Samuel 2005). Some of these criticisms were based on the understanding that one of the most robust and important single predictors of personality pathology is generalized severity of functioning (Hopwood et al. 2011) and the belief that a consideration of severity is essential to any dimensional model of personality (Parker et al. 2002; Tyrer 2005). Accordingly, in the development of DSM-5, the Personality and Personality Disorder (P&PD) Work Group proposed that personality pathology be characterized by impairments in personality functioning, along with the presence of pathological personality traits. Literature reviews (e.g., Bender, Morey, and Skodol 2011) and empirical investigations (e.g., Morey et al. 2011) were completed to identify and support the specific elements fundamental to an individual’s mental representations of self and others, and those central to personality functioning. In spite of substantial empirical support, the APA Board of Trustees voted not to accept the P&PD Work Group’s proposal for reconceptualizing personality disorders, but agreed to include the AMPD presented in Section III of DSM-5 as an emerging model.

One goal of the P&PD Work Group in the creation of the AMPD was to delineate the boundaries between manifestations of personality pathology and subsequent consequences (Narrow and Kuhl 2011), which resulted in the differentiation of disability and dysfunction (Skodol 2012). Because disability, the consequential impairments in functioning, differs from the core disturbances in functioning believed to underlie personality pathology, the AMPD regards personality as characterized by maladaptive personality traits and impairments—or lack of impairment—in personality functioning (i.e., impairments in self- and interpersonal functioning), extending beyond observable symptoms. Thus, according to the AMPD and in accordance with psychoanalytic theory, personality is dimensional; a personality disorder diagnosis is to be assigned when an individual maintains a pattern of impairment in personality functioning (Criterion A) and one or several maladaptive traits persist (Criterion B).

Criterion A is met when an individual experiences moderate or greater impairment in personality functioning, as manifested by difficulties in two or more of the four elements comprising self- and interpersonal functioning, which maximizes the ability of clinicians to accurately and efficiently identify personality pathology. This emphasis on personality functioning brings a new domain for conceptualizing personality and personality pathology to the DSM diagnostic system, but the prominence of self- and interpersonal functioning is hardly a novel development in personality theory or classification. In fact, self- and interpersonal functioning are two of the relatively persistent themes across several theories of personality. These constructs have been identified by myriad scholars, though often using different labels. The constructs have been studied by psychoanalysts as achievement or power and affiliation or intimacy (McAdams 1985; McClelland 1985; Winter 1973; see also Freud 1930), identity and intimacy (Erikson 1963, 1968), anaclitic and introjective configurations (Blatt 1974, 1991, 1995), and innumerable others. Several other schools of personality theory have similarly identified interpersonal and intrapersonal elements to be core dimensions of personality (see Blatt 2008; Luyten and Blatt 2011). Encouragingly, even neurobiological and genetic research has begun to investigate distinct substructures of self-functioning (Dimaggio et al. 2008; Lieberman 2007; Lombardo et al. 2007, 2010; Uddin 2007; for a meta-analysis, see Northoff et al. 2006) and interpersonal functioning (Donaldson and Young 2008, Gordon et al. 2008; Heinrichs and Domes 2008; Levine et al. 2007), as well as the interaction between the two (Vanderwal et al. 2009). The dimensionality of personality and the focus on self and on self in relation to others, nevertheless, is inherently psychoanalytic (Bornstein, Maracic, and Natoli 2018).

Criterion B of the AMPD requires the existence of one, or even several, maladaptive traits. Assignment of specific maladaptive personality traits is based on personality trait theory and the five-factor model (see Costa and McCrae 1992; Costa and Widiger 2002; McCrae and Costa 2008), which theorizes that personality can be reflected in variations of five higher-order traits: Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. In this model, the core components of one’s personality system are designated as basic traits, characteristic adaptations (intrapsychic and interpersonal features that develop), and the self-concept. It has been suggested, rightly so, that the five-factor model was influenced by and built on earlier personality theories, some of them psychoanalytic (Digman 1990).

A supplemental classification system of Personality and its comparison with the two diagnostic models of dsm-5

The PDM presents a dimensional understanding of personality through a psychoanalytic lens, and offers a framework covering the full range and depth of mental functioning that describes the dynamics underlying different character organizations. Specifically, the diagnostic framework of the PDM incorporates an individual’s personality syndromes (the P Axis), his or her unique profile of mental functioning (the M Axis—patterns of relating, approaches to coping with stress and anxiety, style of comprehending and expressing feelings, capacity for mentalization, and ability to form moral judgments), and one’s personal, subjective experiences of his or her symptoms (the S Axis). The PDM system combines these with descriptions of the individual’s level of personality organization (i.e., a dimensional view of the severity of personality impairment, from healthy personality to severe personality disorder) and personality patterns. In 2017 the Alliance of Psychoanalytic Organizations released the second edition of the PDM, which includes several changes and innovations. Among these innovations are the inclusion of a psychotic level of personality organization and a general expansion of the descriptions of each pattern of personality. The PDM’s original approach to conceptualizing and classifying personality and personality dysfunction remains similar across the two editions.

Comparison of the PDM’s framework for classifying personality, the DSM-5’s categorical model, and the DSM-5’s AMPD reveals a number of dissimilarities, as well as many areas in which these systems converge. An important distinction among these systems is whether the formulation of personality pathology is categorical, whether it can be appropriately executed on the basis of specific, often observable traits, or whether attention must also be given to intrapsychic themes (McWilliams 2012). As reviewed above, DSM-5’s categorical model guides clinicians to assign a personality disorder diagnosis when an individual maintains a set number of pervasive symptoms that result in clinically significant distress or impairment. Several criticisms have been made of this categorical model (Clark, Livesley, and Morey 1997; Cloninger 2000; Livesley 2003; Oldham and Skodol 2000; Rounsaville et al. 2002; Skodol 2014; Tyrer 2001; Widiger 1993; Widiger and Samuel 2005). One objective in developing the AMPD was to address these concerns, particularly those which would be reconciled through a decrease of comorbidity (Skodol 2012). And although the AMPD describes various thoughts and behaviors associated with each trait, some have argued that this new approach still fails to offer explanations of how different traits might be connected when present within a single individual or in what context the associated behaviors might manifest (McWilliams 2012). In contrast, PDM organizes personality around underlying dynamics and core themes (e.g., areas of preoccupation, tension, and/or conflict) in addition to the consideration of self- and interpersonal functioning.

A second key difference between the personality models of DSM-5 and PDM-2 is the attention paid to a full spectrum of personality, from healthy, adaptive functioning to extreme impairment. DSM-5’s categorical approach to conceptualizing personality, and mental health in general, is almost completely void of any recognition of healthy personality. Inclusion of the World Health Organization’s Disability Assessment Schedule, version 2.0 (see Üstün et al. 2010), permits some acknowledgment of adaptive functioning, but only in terms of absence of impairment. More effectively, the AMPD directly recognizes that an optimally functioning individual has a complex, fully elaborated, and well-integrated psychological world that includes a mostly positive, volitional, and adaptive self-concept; a rich, broad, and appropriately regulated emotional life; and the capacity to behave as a productive member of society with reciprocal and fulfilling interpersonal relationships (DSM-5, p. 771).

Moreover, the Level of Personality Functioning Scale in the AMPD can be used as a global measure of personality functioning. The Level of Personality Functioning Scale, which serves as a measure of an individual’s current overall level of self- and interpersonal functioning impairment, affords clinicians the ability to identify subthreshold and healthy areas of personality functioning. Though an improvement in accounting for healthy, adaptive functioning over former approaches, this minor inclusion is a far cry from the level of attention the PDM pays to the full continuum of personality. True to psychoanalytic tradition, the PDM makes a clear effort to structure its approach to conceptualizing personality as a method for understanding an individual’s overall psychological makeup within the context of the whole person. To that end, the PDM identifies the distinction between a personality type or style and a personality disorder, extending its usefulness beyond that of the DSM by presenting prototypes of these various personality styles and discussing the processes taking place across the different levels of personality organization.

DSM-5’s categorical model, DSM-5’s AMPD, and PDM-2 converge on multiple points (see PDM-2 for a comparison of PDM-2 with other diagnostic systems). One convergence in particular is the overlap between many of the personality patterns detailed in PDM-2 and the specific personality disorders found in the AMPD and the categorical model of DSM-5 (e.g., PDM’s anxious-avoidant and phobic personality and DSM’s avoidant personality disorder). There are a few exceptions (DSM’s schizotypal personality disorder, DSM’s borderline personality disorder and ICD-10’s emotionally unstable personality disorder, and the DSM/PDM narcissistic personality disorder), and the PDM identifies additional personality patterns—sadistic and sadomasochistic, dissociative, masochistic (self-defeating), depressive, somatizing, and anxious. There are, however, more striking similarities observed between PDM and the AMPD specifically. Excitingly, an explicit similarity between PDM and the AMPD reveals a clear connection of the DSM system back to psychoanalytic theory.

The DSM’s “Hidden” Reconnection to Psychoanalytic Theory

One of the major organizing principles of PDM-2’s P Axis is level of personality organization, a second being personality style or type. These descriptions of an individual’s level of personality organization (i.e., a dimensional view of the severity of personality functioning impairment) closely correspond with the DSM-5 P&PD Work Group’s goal of separating clinical severity from personality expression in diagnosis. This goal was achieved, consciously or unconsciously, using the psychoanalytic principles that serve as a foundation for Criterion A of the AMPD (Hopwood et al. 2011). Although there is similarity between Criterion A and Criterion B of the AMPD and these two features of PDM-2’s P Axis, it is the centrality of self- and interpersonal functioning in both systems that reveals DSM’s reconnection to psychoanalytic theory. Specifically, the proposal by the DSM-5 P&PD Work Group, which became Criterion A of the AMPD, dictated that “disturbances in self- and interpersonal functioning [i.e., personality functioning] constitute the core of personality psychopathology” (DSM-5, p. 762). This recognition strongly resembles central tenets seen in attachment theory and object relations literature, two prominent fields of psychoanalytic theory that have heavily influenced the PDM system.

Despite numerous variations of object relations theory that have emerged over the years (see Greenberg and Mitchell 1983), object relations can be broadly defined as an individual’s mental representations of self and important others (e.g., parents, siblings), including the wishes, emotions, fantasies, and fears attached to these internal representations. Noticeably, internalized mental representations of self and important others (sometimes called introjects) can be compared to self-functioning and interpersonal functioning, respectively. To further substantiate this parallel, let us more closely consider the presentation of self- and interpersonal functioning in one of the most prominent object relations models of personality today, Blatt’s anaclitic-introjective framework (1974, 1991, 1995; see also Besser, Guez, and Priel 2008; Blatt 2008).

Interpersonal functioning and functioning of the self can be recognized within Blatt’s theory as being similar to the anaclitic and introjective configurations of personality (see Blatt and Blass 1996; Blatt, Shahar, and Zuroff 2001; Blatt and Shichman 1983). Blatt’s model of personality development and psychopathology proposes that one’s personality develops along two essential developmental lines—an anaclitic line and an introjective line. The anaclitic (or relatedness) line “involves the development of the capacity to establish mature, mutually satisfying interpersonal relationships” (Blatt, Shahar, and Zuroff 2001, p. 499) and can be likened to the domain of interpersonal functioning. The second developmental line in this model, the introjective (or self-definitional), is equivalent to the domain of self-functioning and “involves the development of a consolidated, realistic, essentially positive, differentiated and integrated self-identity” (Blatt, Shahar, and Zuroff 2001, p. 499). This distinction has been demonstrated to serve an important role in the conceptualization of diverse forms of psychopathology and has shown great potential to inform research and practice (Blatt 1974; Blatt and Auerbach 1988).

Attachment theory incorporates constructs similar to those of self- and interpersonal functioning. It maintains that many facets of personality, including intra- and interpersonal characteristics, are influenced by variability in one’s early attachment to caregivers (Bowlby 1969, 1973, 1977). Akin to the introjects of object relations theory, attachment theory’s internal working models—a cognitive framework consisting of mental representations of self and attachment figures—become important features of each individual’s unique personality structure (Bartholomew and Horowitz 1991; Bowlby 1973; Bretherton and Munholland 1999). As hypothesized by Bowlby (1969), these internal working models assist the individual in understanding his or her experiences, in terms both of interpersonal interactions and of the self. Self- and interpersonal functioning can further be likened to attachment avoidance and attachment anxiety, two underlying constructs contributing to one’s attachment style and personality. As a whole, the way in which one perceives self and others, and the relative contrast of these perceptions, are influential components in attachment theory’s conceptualization of personality (Levy, Johnson, et al. 2015). The association between personality pathology and impairment in self- and/or interpersonal functioning has been acknowledged (or alluded to) in the work of many attachment theorists and researchers (e.g., Fonagy 1999; Fonagy et al. 1995; Gunderson 1996; Levy and Blatt 1999; Levy, Scala, et al. 2015; Meyer and Pilkonis 2005).

Conclusion, Implications, and Moving Forward

Through the long history of personality theory, the initial standardized systems of personality classification and diagnosis were influenced by the knowledge and models that preceded their development. The DSM in particular was once heavily informed by psychoanalytic theory (Shorter 2005). However, the APA’s desire for greater neutrality with respect to etiological theories encouraged a shift away from psychoanalytic theory, resulting in the problematic atheoretical model introduced in DSM-III in 1980. Despite an ongoing emphasis on the use of noninferential personality disorder categories, the claim made in DSM-5’s AMPD that “disturbances in self and interpersonal functioning constitute the core of personality psychopathology” (p. 762) directly parallels primary themes that characterize psychoanalytic personality theories (Bornstein, Maracic, and Natoli 2018). As demonstrated by the comparison with PDM-2 within the context of two prominent psychoanalytic theories, the AMPD brings psychoanalytic theory back into the DSM system.

There are multiple implications for a more psychoanalytic DSM. Foremost is the increased clinical utility of the dimensional AMPD over the DSM categorical model, seemingly as a result of its inclusion of core psychoanalytic principles and practices. Psychoanalysis and psychodynamic psychotherapies often place an emphasis on considering each patient’s level of personality organization regardless of treatment goals or diagnosis, as a patient’s location on this dimension will guide case formulation and clinical treatment (McWilliams 2004, 2011). Paralleling this practice, Criterion A of the AMPD not only permits, but encourages, an evaluation of each patient’s level of personality functioning. This feature of the AMPD gives clinicians the opportunity to formulate a greater understanding of each patient’s overall psychological makeup than was possible with earlier editions of the DSM. Second, the AMPD’s shift toward the psychoanalytic perspective of personality has potential to increase the dialogue between psychoanalytic theory and other clinical theories (e.g., cognitive-behavioral) and offers a common language with which psychoanalysis can deepen its links with other mental health and medical disciplines. However, there is still a significant gap between DSM-5 and psychoanalytic models of personality (Vanheule 2017). Moving forward it will be necessary to provide clinicians a comprehensive and clinically applicable method for conceptualizing each patient’s personality style or type in addition to his or her level of personality functioning impairment. Although the AMPD currently offers a trait-based framework for describing a patient’s character, the future inclusion of psychodynamically informed case descriptions may prove advantageous, as research has indicated that clinicians prefer these over trait-based descriptions like those in the AMPD (Spitzer et al. 2008). That is, supplementing the AMPD’s personality trait model (i.e., Criterion B) with a psychodynamically informed assessment instrument such as the Shedler-Westen Assessment Procedure (SWAP; Westen and Shedler 1999a,b) may be the next step in increasing the clinical utility of DSM by reconnecting it to psychoanalytic theory.