Integrating and differentiating personality and psychopathology: A psychodynamic perspective

Several strands of research converge to suggest that personality and psychopathology can be integrated in the form of a hierarchical model of individual differences. The notion that personality and psychopathology are intrinsically linked has a long tradition within psychodynamic approaches. In this article, we first summarize empirical evidence supporting two related key assumptions of psychodynamic approaches to personality and psychology: that a developmental, person-centered approach

general categories that were clinically highly popular, such as neuroses, disappeared in the forest of specific disorders that came to dominate descriptive psychiatry.A brave few, like Peter Tyrer (1985Tyrer ( , 1990) ) and some others (Andrews et al., 1990;Goldberg, 2015), stood up to this trend.
The developers of DSM had hoped that a descriptive approach to classification would foster research on the underlying principles of psychiatric categorization, thus enabling the development of a more empirically based classification system in the future.Hence, the basic assumption was that psychiatric disorders could be largely distinguished based only on objectively assessable symptom criteria.It soon became apparent that this approach was more difficult than anticipated.Not only did it soon become clear that etiological considerations had to be taken into account in classifying some psychiatric disorders (e.g., mood disorders due to a general medical condition); it also became quickly apparent that personality features needed to be included as "descriptive features" of psychiatric disorders-not only for personality disorders but also for many "symptom" disorders (e.g., in delineating dysthymic disorder) (Westen & Weinberger, 2004).More generally, studies increasingly suggested that the association between Axis I (symptom disorders) and Axis II (personality disorders) was more intrinsic than assumed (Luyten & Blatt, 2011;Westen & Weinberger, 2004).
Perhaps the greatest source of dissatisfaction with diagnostic taxonomies arises from the high frequency of comorbidity (Kessler et al., 2005;Merikangas et al., 2010), as has been repeatedly demonstrated in large-scale surveys in the USA (Kessler et al., 2005(Kessler et al., , 2011) ) and the UK (Bebbington et al., 2009;Farrell et al., 2003), and is common in both clinical and community samples (e.g., Barlow et al., 2014;Cummings et al., 2014;Ormel et al., 2015).Comorbidity transcends diagnoses, and there is overlap between most symptoms that are known to be associated with mental disorders (e.g., Budde et al., 2018).In fact, the big change came not in 1980, when DSM-III was introduced, but in 1997, with the publication of DSM-III-R (American Psychiatric Association, 1987), when the exclusion criteria used to implement diagnostic hierarchies were abolished (Beauchaine & Cicchetti, 2016).The hierarchy was due for demolition because it was not empirically based but embodied a subjective rank ordering of seriousness from a psychiatric standpoint.Its abolition led to a mushrooming of comorbidities, along with research on comorbidity (Angold et al., 1999).This early research was descriptive and raised questions in relation to the utility of the system of taxonomy, but not questions of mechanism, causation, or treatment.A brief glance at PubMed tells the story: in 1990, there were 171 articles that mentioned comorbidity; by 2000 there were 2,040, and by 2010 there were 7,898.
Notwithstanding acute problems with a taxonomic approach, for several decades, other approaches to psychiatric classification faded to the background, at least in mainstream psychiatry and psychology (Blatt & Luyten, 2010;Shedler et al., 2010), as their basic assumptions about the classification and diagnosis of mental disorders were often markedly different from the dominant descriptive approach.This was definitely the case for psychodynamic approaches to psychiatric classification.These approaches are fundamentally developmental and person-centered; they focus on developmental pathways from infancy to adulthood and on how disruptions in developmental tasks as a result of complex interactions among biological and psychosocial factors are expressed across the lifetime (Lingiardi & McWilliams, 2017;Luyten & Blatt, 2011;Luyten et al., 2012;McWilliams, 2011).Hence, rather than focusing on discrete disorders, these approaches focus on how individuals negotiate different developmental tasks and how problems in negotiating these tasks may be expressed in different ways across the lifespan.Thus, psychodynamic approaches to psychiatric classification are fundamentally focused on providing a "taxonomy of people" rather than a "taxonomy of diseases" (Lingiardi & McWilliams, 2017, p. 17).Stated otherwise, they aim to describe "what one is rather than what one has" (Lingiardi & McWilliams, 2017, p. 17).This view contrasts with the static and largely nondevelopmental approach in DSM and in similar descriptive approaches to psychiatric classification.Moreover, from a psychodynamic perspective, there is no neat distinction (at least for most psychological problems) between "normality" and "pathology".Moreover, psychological "disorders" are seen not as static end states, but as functional-that is, as attempts at adaptation.Although in the long run these attempts at adaptation might become highly maladaptive for the individual and his or her environment, they are thought to represent the best possible balance for the individual between his or her psychological endowment, biological predisposition, and environment.
This recently newly revived functional approach to psychopathology also centrally emphasizes the role of environmental factors in the etiology of psychopathology.Once again, this approach markedly contrasts with the descriptive DSM approach, which, with notable exceptions (e.g., posttraumatic stress disorder) attempts to define psychiatric disorders without much consideration of environmental factors as a result of the symptom count/cut-off method that is at its core.Perhaps less sharply contrasting is the emphasis of psychodynamic approaches to classification on biological factors in the development of psychological problems.Within the psychoanalytic tradition, there has always been a recognition of the role of biological factors in psychopathology, and hence of the idea that a disorder-centered "disease" model is needed to complement a developmental, person-centered approach.Similarly, psychoanalytic researchers and clinicians have always recognized the value of diagnostic labels (and, indeed, have historically been instrumental in delineating several types of psychopathology, such as obsessive-compulsive disorder and borderline personality disorder).As a result, psychodynamic approaches to psychiatric classification have always combined a descriptive, disorder-oriented approach with a developmental, person-centred, etiologically based approach.Last but not least, psychodynamic approaches toward classification also led to a quite different view on the prevention and treatment of psychological disorders.Whereas the disorder-centered approach promoted by DSM led to the development and empirical evaluation of a wide array of treatments for specific disorders, psychodynamic approaches have always embraced a more transdiagnostic view of psychological treatment that emphasizes broad-spectrum rather than disorder-specific treatments (Fonagy et al., 2006).
As evidenced by the papers in this Special Section, recent years have seen a renewed interest in more theoretically based approaches to psychopathology rooted in theories about normal and disrupted personality development, not only in terms of the conceptualization and classification of psychopathology, but also with regard to the treatment of psychological disorders.With the increasing realization that psychological disorders often share common mechanisms, there has been a renewed interest in transdiagnostic interventions (Barlow et al., 2014;Weisz et al., 2012).This realization has also led to a resurgence of interest in psychodynamic approaches to the conceptualization and treatment of psychological disorders (Blatt & Luyten, 2010), most clearly demonstrated by the influence of psychodynamic formations concerning the centrality of different levels of impairments in self and relatedness in classifying and diagnosing personality disorders in DSM-5 (American Psychiatric Association, 2013) in an attempt to improve the reliability and validity of personality disorder classification (Skodol, 2012).Yet, even further, personality-based models of psychiatric classification-in particular, the Hierarchical Taxonomy of Psychopathology (HiTOP) approach (Kotov et al., 2020;Ruggero et al., 2019), which take an analogous approach to psychodynamic thinking, have been instrumental in questioning some fundamental assumptions of the descriptive approaches to psychiatric classification and the distinction between symptoms, disorders, personality, and personality disorders as such.
In this article, we summarize the main assumptions of psychodynamic approaches to psychiatric classification and the empirical evidence supporting these assumptions, with a focus on the relationship between personality and psychopathology.With the growing popularity of personality-based approaches to psychopathology as an alternative to the current descriptive approach to the classification of psychiatric disorders, the long-standing question of whether we can separate the person from the disorder is back at center stage in discussions concerning the future of psychiatric classification.To answer this vexing question, we first need to clarify what we understand by the concepts "person", "personality", and "(personality) disorder".Here, we propose a new conceptualization of both personality and psychological disorder based on a socialcommunicative approach to personality development, and discuss the implications of these views for the ongoing discussion about the extent to which person and disorder can be separated.We will approach this question from two angles.First, we will discuss psychodynamic approaches to the understanding of psychological disorders, and their implications for discussions concerning the distinction between personality and psychological disorders.We then focus on the concept of personality and outline a newly emerging understanding of personality and psychological disorders.We close this article with a discussion of the implications of these views for ongoing research on the nature of psychiatric disorders and their treatment.

| THE NATURE OF PSYCHOLOGICAL DISORDERS
As in many areas of psychology and psychiatry, the science of psychiatric classification is faced with a major problem, in that there is no objective standard to which theoretically or empirically derived classification approaches can be compared.Hence, as Gödel (1933) proposed in mathematics, a higher order of abstraction is needed to test the validity of a particular classification system, as some of the basic propositions of a given approach to psychiatric classification cannot be empirically tested, nor is it possible to demonstrate that the basic propositions of a given classification system are consistent precisely because of the absence of an objective standard.Psychodynamic approaches attempt to solve this problem by delineating empirically supported key principles to assess the validity of any given approach to psychiatric classification.
In the context of this article, the following two basic assumptions underpinning psychodynamic approaches to psychiatric classification that bear on our understanding of psychological disorders can be distinguished: a.A person-centered, developmental psychopathology approach is essential in psychiatric classification b.Functionality: psychopathology typically represents an attempt at adaptation We discuss each of these assumptions and supporting empirical research in the following sections.

| A person-centered, developmental psychopathology approach is essential in psychiatric classification
Although there is much still to be discovered about the developmental origins of psychological disorders, research findings have demonstrated the complex, multifactorial, and often recursive causality involved in normal and abnormal human development throughout the lifespan (Caspi et al., 2016;Cicchetti, 2016).Developmental pathways implicated in vulnerability to psychopathology seem to be centrally governed by equifinality and multifinality (Cicchetti & Rogosch, 1996): different etiological factors are implicated in similar developmental outcomes (equifinality), whereas, depending on interactions with other factors, a specific etiological factor (e.g., temperament) may lead to different developmental outcomes (multifinality).From the perspective of developmental psychopathology, comorbidity may be considered largely artefactual, due to nosology splitting disease entities into sub-categories that likely lack validity (Beauchaine & Cicchetti, 2016;Forbes et al., 2016).If disorders were reconceptualized as indicators of latent transdiagnostic spectra, comorbidity would no longer be a problem.If disorders share phenotypic and genotypic variance, "comorbidity" seems to be an inappropriate term, which could be readily replaced by the phrase "frequent co-occurrence" (Goldberg, 2015).The term "comorbidity" should be reserved for etiologically distinct disorders-such as, for example, bipolar disorder and hiatus hernia-as was originally intended (Feinstein, 1970).
Adopting a developmental psychopathology perspective first and foremost implies that psychiatric classification should be rooted in longitudinal developmental research (Caspi et al., 2014).Specifically, classification efforts should be based on the consideration of developmental pathways that lead to a range of psychological states across the lifespan.From this perspective, for the vast majority of developmental outcomes, it seems difficult to distinguish the person from the disorder, as both seem to be the outcome of a complex set of factors that shape psychological development (Caspi et al., 2014).This also raises the question of whether there is a neat distinction between "normal" and "disrupted" personality development, and between "normality" and "pathology" more generally, consistent with research findings suggesting that most psychological variables, including personality and psychological disorders, are best represented on a continuum and thus do not represent taxons.For instance, a recent meta-analysis of 317 findings drawn from 183 papers found overwhelming support for a dimensional view of psychopathology (Haslam et al., 2020); only a few psychological variables (e.g., autism and alcohol use) emerged as candidate taxons.
A neat distinction between a person and his or her disorder is thus not supported by research on vulnerability factors implicated in psychological problems.Basic evolutionary-based psychobiological systems are implicated in the development of both "disorders" and "personality".For instance, impairments in the stress and mesocorticolimbic reward system are implicated in the development of both broad-bandwidth (e.g., neuroticism) and narrow-bandwidth (e.g., self-critical perfectionism) personality dimensions, as well a wide variety of psychological disorders, ranging from depression to conduct disorder and personality disorders (Fonagy & Luyten, 2018;Lupien et al., 2009;Luyten & Fonagy, 2018).Hence, categorical, descriptive approaches to psychiatric classification seem to have resulted in the reification of "disorders".Something similar seems to have happened with regard to the notion of "symptoms".Although some features of psychopathology may be similar to symptoms that define a medical disease (e.g., disturbances in circadian rhythm in depression, or hallucinations in psychosis), the vast majority of "symptoms" of psychological disorders as defined in DSM and other classification systems refer to personality functioning (e.g., guilt, low self-esteem, irritability, perfectionism, or narcissism).Hence, more care is needed in distinguishing between symptoms in the medical sense of the term and personality features defining psychological disorders.Otherwise, not only is confusion created in delineating and diagnosing psychological disorders, but there is also the risk of circularity.For instance, what does it mean to say that neuroticism predicts severity of depression if features of neuroticism such as guilt and low self-esteem are included in the measure of severity of depression (i.e., the dependent variable)?Although this statement superficially appears to follow the same logic as to say that "a bacterium causes an infection and fever," in reality it does not.
The temporal stability of most common mental health problems similarly points to problems in distinguishing the person from the disorder.The expectation that the former is more stable across time than the latter is less robust than one might think.For instance, estimates of the temporal stability of depression (r = 0.50-0.60)are similar to the estimates of personality traits assumed to predict depression (Caspi et al., 2005).
The finding that historical and sociocultural factors influence judgements about normality and pathology in psychological functioning similarly points to problems distinguishing the person from the disorder (Stevanovic et al., 2017).Where does "normal variation in personality" end and "psychological disorders" begin?Despite the fact that within a given culture there is often considerable consensus as to what psychological states or behaviors can be considered to reflect a psychological disorder (e.g., referring to negative effects of a person's behavior for him/herself and/or others), this consensus is rarely absolute-and, even if there is a consensus, the boundary between what is considered normal and pathological is still quite blurred.
Moreover, the majority of mental disorders have their first onset in childhood and/or adolescence (Kim-Cohen et al., 2003), and children and adults typically present with multiple problems, either cross-sectionally or sequentially across the lifespan (Caspi et al., 2014).Hence, as noted, the high observed comorbidity between psychological disorders | 5

LUYTEN aNd FONaGY
seems to be an artifact of a categorical descriptive classification system rather than reflecting true comorbidity.Even in children with seemingly isolated and relatively mild psychological problems, these problems might be indicative of a broader underlying vulnerability.For instance, in a population-representative study as part of the World Health Organization's World Mental Health surveys, among 123,628 respondents from 25 countries the lifetime prevalence of internalizing psychological disorders increased from 18.2% among those without a childhood phobia to 46.3% among those with one subtype of childhood phobia and 75.6% among those with more than four subtypes (de Vries et al., 2019).Moreover, while the rate of lifetime suicide attempts in those without a history of childhood phobia was low (2.0%), it was higher in those with one subtype of childhood phobia (6.5%) and reached 16.8% among those with more than four subtypes.Hence, childhood phobia was related to adverse outcomes throughout the life course, suggesting that it may often be indicative of a broader underlying vulnerability.Hence, consistent with psychodynamic approaches to psychopathology, "comorbidity" is the norm rather than the exception and points to an underlying vulnerability that is expressed at different life stages in different ways.Such sequential comorbidity is the rule rather than the exception.
Take, for instance, the example of David, who since the age of 12 has had the strong feeling that his parents have always preferred his sister because she is more intelligent and attractive than him.As a result, David developed very negative self-views, while at the same time feeling frustrated and angry with his sister and parents because of the way they have made him feel.Yet, because he also loved his sister and parents and admired them, he felt it difficult to acknowledge his negative feelings.Instead, it was safer for him to be angry with himself and with other adults whom he perceived to be in a position of authority, such as teachers and the police.As a result, and as peer pressure to assert himself as a strong and attractive young man increased, he increasingly felt that he could not live up to what was expected from him.In an attempt to assert his autonomy and self-worth, David began to behave in ways that were quickly identified by his teachers and parents as "oppositional" behavior (e.g., he often argued with teachers and refused to do his homework).In an attempt to be considered one of the "cool kids", he became involved in a group of teenagers who experimented with alcohol and cannabis.David also felt very attracted to one of his classmates; they become romantically involved, but she quickly ended the relationship because she found him to be too possessive and controlling.David felt rejected, abandoned, and hurt; his externalizing behavior increased in an attempt to relieve feelings of frustration and aggression, which he directed sometimes to himself (e.g., he engaged in self-harming behavior, such as hitting the wall with his fist until it bled) and sometimes to others (e.g., he became a bully to his classmates).When he was in his early twenties, he continued to bully his peers (now his colleagues at work), continued to struggle with feelings of depression and anxiety, self-medicated with alcohol, and often arrived at his workplace drunk.Finally, after a severe alcohol intoxication, he was admitted to a psychiatric hospital, where he was diagnosed with depression and narcissistic, antisocial, and borderline personality disorder.Yet, it must be clear that David does not in fact have six (or more!) different "disorders" (i.e., depression, oppositional defiant behavior, substance abuse disorder, narcissistic personality disorder, antisocial personality disorder, and borderline personality disorder), but that he has struggled with issues related to selfworth and self-esteem that are expressed at different times and in different ways throughout his life.
Psychological disorders are best not conceptualized as static end states, but as complex, multidetermined, dynamic conflict-defense constellations.This brings us to the view that psychopathology is functional.

| Functionality of psychopathology
It has always been a central assumption of psychodynamic approaches to psychiatric classification that most forms of psychopathology should be seen as distorted modes of adaptation that reflect variations in personality development (Blatt, 2008;McWilliams, 2011).From this perspective it is again difficult to separate the "person" from the "disorder".
Empirical research has provided ample support for the notion that different types of personality and psychopathology reflect attempts at adaptation.For instance, there is increasing consensus that specific attachment styles reflect adaptations to a particular environment (Ein-Dor et al., 2010;Ellis et al., 2011).Attachment deactivating strategies-which are observed in, for instance, fearful-avoidant and dismissive attachment-typically represent a response to the (perceived) unavailability of attachment figures early in life.Individuals who engage in these strategies have learned from these early experiences not to rely on others; even when in need, they have the (implicit) belief that they have to be able to deal with problems on their own, often leading to compulsive autonomy.Individuals who primarily rely on attachment hyperactivating strategies (as observed in those with preoccupied attachment) characteristically have grown up in an environment characterized by inconsistently available attachment figures, and thus had to resort to often drastic strategies to elicit care and support from their attachment figures.
Experimental studies have provided further evidence for the view that these so-called "insecure" attachment strategies present conflict-defense constellations.Individuals who primarily rely on attachment deactivating strategies, for instance, often present a positive model of the self and show a tendency to deny distress, but both priming and physiological studies suggest that this is a defensive strategy to ward off underlying feelings of distress, insecurity, and inferiority (Mikulincer & Shaver, 2007).Similarly, individuals who primarily rely on attachment hyperactivating strategies typically report having positive models of others, but they also harbor feelings of jealousy and anger toward others, which they typically inhibit due to fear of abandonment and rejection (Mikulincer & Shaver, 2007).
Furthermore, studies suggest that when attachment figures are the source of both love and support but also abuse and neglect, the child typically becomes trapped in an approachavoidance conflict, leading to disorganization of the attachment system (Holmes, 2004;Main & Hesse, 1990) expressed in an oscillation between attachment hyperactivating and deactivating strategies.Consistently, studies suggest individuals with complex trauma and disorganized attachment hold views of both the self and others that oscillate between extremely positive and attractive, and extremely negative and repulsive (Luyten, Campbell, & Fonagy, 2020).
There is now also good evidence to suggest that, like personality, most types of psychopathology are equally functional, as they represent attempts to find a balance, however distorted, between internal and external exigencies.For instance, patients with obsessive-compulsive disorder exhibit high levels of both conscientiousness and latent impulsive aggression that is experienced as ego-dystonic; thus, obsessive-compulsive symptoms seem to reflect strategies to cope with high levels of impulsive aggression (Moritz et al., 2011).Patients with narcissistic personality disorder typically present with high self-esteem, which serves to compensate for underlying feelings of inferiority and emptiness (Russ et al., 2008).Hence, these patients simultaneously exhibit "high" and "low" levels of self-esteem; this is difficult to explain with multivariate trait models that assume that individuals are either high or low on a particular trait, but not both.Yet, contrasting and conflicting feelings seem to be the hallmark of both normal and disrupted personality development (and thus psychopathology).
Likewise, patients with dependent personality features, which are implicated in a wide range of disorders, from depression and anxiety to borderline personality disorder (Luyten & Blatt, 2013), typically exhibit low levels of manifest aggression but high levels of latent aggression, which is only indirectly manifested (e.g., in passive-aggressiveness).Individuals with avoidant personality disorder features simultaneously desire and fear intimacy (Westen et al., 2003).Hence, a consideration of the functional or motivational features of both personality and psychopathology (i.e., as attempts to defend against or cope with both internal and external exigencies) is essential.This leads to the view that both personality and psychological disorders are dynamic conflict-defense constellations, rather than static end states.We will return to this view in more detail in the next section.
First, however, we want to make it clear that the findings summarized above do not invalidate multivariate trait approaches (Luyten & Blatt, 2011).On the contrary, there is now increasing evidence that basic, broad-bandwidth personality factors may play an important role in determining the specific developmental trajectories of personality development (and thus psychopathology).Using Waddington's (1957) famous epigenetic landscape metaphor, basic temperament or personality traits identified by multivariate models may nudge or even knock personality development in a particular direction in the epigenetic landscape that determines developmental pathways.For instance, considering the relationship between self-criticism and depression, there is ample evidence to suggest that self-criticism is associated with increased vulnerability for depression and anxiety, even when controlling for broad-bandwidth traits such as neuroticism (Kopala-Sibley et al., 2017).Furthermore, self-criticism has been shown to reflect attempts to prove one's self-worth and lovability in individuals whose caregivers were emotionally unavailable and/or provided love and care only conditionally (Vandenkerckhove et al., 2020).Empirical research in this context suggests, for instance, that adolescents with low levels of effortful control tend to use externalization as a way of coping with self-critical issues (e.g., by engaging in antisocial behavior directed toward authority figures).By contrast, adolescents with higher levels of effortful control (disinhibition) tend to blame themselves and thus are at increased risk for developing internalizing symptoms such as self-critical depression (Leadbeater et al., 1999).Similarly, temperamental factors, such as high levels of aggression, have been found to moderate the impact of early adversity and inconsistent parenting in particular on the development of disorganized attachment and associated psychological disorders such as borderline personality disorder (Luyten, Campbell, & Fonagy, 2020).

| PERSONALIT Y AND PERSONALIT Y TRAITS
The views and findings reviewed above have important implications for our understanding of what "personality" is and what it is not.The concept of personality usually refers to the view that people have characteristic patterns of thoughts, feelings, and behaviors that are consistent and stable over time and across contexts (Caspi et al., 2005).Like the notion of psychological or psychiatric disorder, the concept of personality is therefore a theoretical construct.In the context of this article, it is key to understand that traditional personality theories tend to attribute the stability and consistency to something in the person.Indeed, traditional personality theories are largely rooted in folk psychology, and personality psychology therefore seems to be an extension of folk psychology, wherein people assume that there are relatively stable differences between individuals in terms of their psychological functioning, how they perceive the world, and how they interact with others.As we will see, recently emerging research findings urge us to reconsider, at least in part, the view that personality (and, by extension, personality disorder) refers to something in the person.
As is well known, there have been several challenges to the view that personality traits are stable across time and contexts, most notably from radical behavioral perspectives (Malone & Cruchon, 2001) and the cognitive-affective processing approach formulated by Mischel and Shoda (1998).Both approaches criticized traditional personality theory, arguing that there is little consistency or stability in personality traits and that behavior is largely under environmental control.Yet, research findings have made it clear that there is very little support for extreme positions in this debate in either direction, and these findings have also been taken aboard in more recent multivariate approaches to personality (Baumert et al., 2017;Bleidorn et al., 2020;Roberts & Robins, 2004).Test-retest correlations suggest moderate absolute stability in personality traits across the lifespan, with correlations increasing with age, ranging from r = 0.41 in adolescence to r = 0.55 by age 30, to approximately r = 0.70 around age 70 (Caspi et al., 2005).Moreover, heritability estimates of personality traits, as for many psychological disorders, are typically in the range of 0.50 ± 0.10 (Caspi et al., 2005;Distel et al., 2008;Torgersen et al., 2012), which could in part explain the relative stability of both personality traits and psychological disorders.Moreover, stability and interindividual differences in "personality" traits have been observed in nonhuman animals (Gosling & John, 1999), suggesting the crossspecies relevance of the concept of personality as something that is relatively stable across time and contexts.Anyone who has owned a pet or is otherwise familiar with animal behavior will appreciate the large individual differences in "personality traits" among animals.
However, although there is considerable stability, there is also considerable room for change in-and thus plasticity of-personality.Estimates of absolute stability are far from perfect.Decreases in rank-order stability in personality as the time between observations increases similarly suggest that personality is malleable (Caspi et al., 2005).Likewise, meta-analyses have found moderate stability in attachment styles across the lifespan, with estimates of stability, as for broad-bandwidth personality traits, increasing with age.Fraley (2002), for instance, reported an asymptotic testretest correlation of r = 0.39 in a meta-analysis of longitudinal studies of attachment from age 1 to 19, whereas the stability of attachment in adulthood increased to r = 0.54 (Fraley & Brumbaugh, 2004).Importantly, risk status (e.g., family conflict, parental separation, minority status, or male sex) has consistently been associated with lower stability of attachment in meta-analyses (Fraley et al., 2011;Pinquart et al., 2013).
Taken together, findings concerning the greater stability of multivariate broad-bandwidth personality traits and attachment styles in adulthood (and particularly in old age) than in childhood and adolescence, and the association of risk status with lesser stability in personality features, makes sense only if personality is conceptualized at least in part in terms of a an adaptation strategy: as the environment changes, so does the individual's attachment style or personality as a result of active and/or passive person-environment correlations.In adulthood, and particularly in old age, there is on average greater stability of the environment, and thus greater stability of personality/attachment (Fonagy et al., 2017b;Fraley, 2019).This assumption is also borne out by findings showing that there is greater stability over time in attachment to parents than in romantic attachment, suggesting a greater need for the continuous revision of internal working models with regard to oneself and others in romantic attachment relationships compared with attachment working models in relation to one's parents (Fraley et al., 2011).Simulation studies point in the same direction: without stability of the environment/context, the stability of attachment is reduced almost to zero (Fraley, 2019;Fraley & Brumbaugh, 2004).The stability of the transgenerational transmission of secure attachment to infants from autonomously attached parents is also moderated by the environmental context (Verhage et al., 2018).Environmental risk undermines transmission: contextual factors appear to hinder parents' ability to tap into their autonomous attachment representations, so that they cannot provide children with experiences conducive to secure attachment relationships.Hence, the potential of parents with autonomous attachment representations to build secure attachment relationships is reduced by the social context.Of course, biological changes in childhood and adolescence, as well as in old age, may facilitate or constrain personality change, as is shown by studies concerning the "forward pull" associated with biological maturation in children and young people and biological decline in old age (Lupien et al., 2009).
But why do some individuals fail to adapt to everchanging circumstances?Most approaches to psychopathology consider rigidity to be the hallmark of psychopathology, whereas "normality" is defined as the capacity to adapt to the inevitable challenges life brings (for a reviews, see Fonagy et al., 2017a).Hence, the question becomes: what explains the often marked absence of flexibility of some individuals?And what does this mean in terms of the distinction between personality and psychological disorder?
In our view, findings concerning a general psychopathology (or "p") factor underlying psychological disorders (Caspi & Moffitt, 2018) are crucially important in this context, and have led to a considerable shift in our views on the nature of personality and the emergence of psychopathology (Fonagy et al., 2017a(Fonagy et al., , 2017b)).Basically, although not uncontested, there is increasing evidence for a superordinate higher order p factor that explains the high correlations between three high-order factors (internalizing, externalizing, and thought disorder) in the structure of psychiatric diagnoses that in turn provide a hierarchical framework explaining the associations between discrete disorders (Caspi et al., 2014;Caspi & Moffitt, 2018).The p factor appears to be a broad transdiagnostic vulnerability factor associated with increased probability of onset of psychological disorders, greater severity and impairment of psychological problems, poorer prognosis, greater familiality, worse developmental histories, and compromised brain function in early life (Del Giudice, 2015;Laceulle et al., 2015;Lahey et al., 2015;Murray et al., 2016).
These views concerning a general severity factor in personality and psychopathology are not new.There has been a long tradition, particularly within the psychodynamic tradition, of arguing that the severity of psychopathology-as reflected, for instance, in the level of personality functioning-is the most important predictor of the course of psychological disorders and their response to treatment (Cierpka et al., 2007;Kernberg & Caligor, 2005;McWilliams, 2011).Traditionally, in these approaches, a continuum of personality pathology is distinguished, ranging from (a) Neurotic level of personality functioning (LPF), characterized by relatively high levels of identity integration, intact reality testing, and the predominant use of more mature (i.e., higher level) defense or coping mechanisms, to (b) Borderline LPF, characterized by identity diffusion and the use of more immature defense mechanisms with relatively intact reality testing, to (c) Psychotic LPF, characterized by marked identity diffusion, impaired reality testing, and the excessive use of immature defense mechanisms.
Figure 1 presents an integration of Caspi's work on the general psychopathology factor with more traditional psychodynamic approaches concerning the level of functioning.In our opinion, this theoretical framework sheds a new light on the relationship between personality and psychological disorders.At the bottom of the pyramid depicted in Figure 1, there is a large group of individuals who present with relatively discrete and somewhat gendered disorders that can be roughly divided into "internalizing" and "externalizing" spectra.At this level, the course and prognosis of psychological disorders are often relatively benign, as indicated by low levels of p factor and thus nonpersistence of psychopathology.Hence, at this level of functioning (i.e., mild neurotic LPF), it may be possible to distinguish not only between relatively discrete disorders (i.e., individuals show low levels of comorbidity) but also between the person and the disorder.When faced with life challenges, many individuals experience psychological problems and distress that are relatively transient and not (yet) engrained in their personality functioning.This assumption is borne out by findings that only one in five people will not have a diagnosable mental health condition at some point in their lifetime (Schaefer et al., 2017) and by findings that the normative response to adversity is so-called minimal impact resilience, implying that most individuals experience only temporary distress when faced with adversity (Bonanno & Diminich, 2013;Southwick et al., 2014).
Yet, in a sizeable proportion of individuals presenting with apparently discrete and mild psychological problems, these problems may be only the tip of the proverbial iceberg, and they may be the first indication of more persistent psychological problems.Earlier, we discussed the example of the relationship between childhood phobia and increases in the lifetime prevalence of internalizing psychological disorders and suicide attempts across the lifespan (de Vries et al., 2019).We also noted the evidence that most psychological disorders have their onset in childhood and adolescence, and that more severe and persistent types of psychopathology are often preceded by apparently mild psychological problems (Caspi et al., 2014).Hence, there is a danger in considering such apparently mild problems as being distinct from the person and his or her developmental history and environmental context, as they may be the first sign of more persistent problems embedded in the individual's personality development in response to a maladaptive context and/or biological vulnerability.As another example, adolescents with clear signs of emerging borderline personality disorder are often misdiagnosed with depression.This leads to high levels of feelings of invalidation, dropout from and lack of response to treatment, and an increasing threshold to seek help for psychological problems.As a result, opportunities for early intervention to try to prevent the full-blown expression of the disorder are missed (Hutsebaut et al., 2020).As we move up the pyramid depicted in Figure 1, psychological problems become more persistent and more engrained in what we may loosely term, using the language of folk psychology, the individual's personality.Moving into borderline LPF, it makes less and less sense to distinguish between discrete psychological disorders.For instance, whereas patients with high-level neurotic LPF may present with depression without marked anxiety or comorbid personality pathology, in patients with borderline LPF feelings of empty depression and anxiety are key defining features of the individual's psychological problems, together with other features (e.g., self-harm, or idealization and denigration of the self and others) that together constitute what we traditionally understand as a personality disorder.Hence, from this perspective, it makes little sense to argue that a given person with borderline LPF has borderline personality disorder with comorbid depression, as intense depression, as part of a general dysregulation of emotional experience, is likely to be a key feature of borderline personality disorder (Luyten, Campbell, & Fonagy, 2020).
At the top of the pyramid, and thus with increasing p factor persistence, the distinction between the person and the disorder becomes even more challenging.Is it productive to try to distinguish between different disorders, as disordered thought processes become the central problem?Here, as often as not, selective responsiveness to medication offers opportunities for validating taxonomic entities such as bipolar disorder and schizophrenia.Research on the influence of "premorbid personality structure" on the course of disease provides a logic for the separation of the personality from the psychopathology (Salaminios et al., 2020).Following the onset of the disorder, "personality" will inevitably be intertwined with the task of adaptation to a severe biobehavioral condition.
What explains the persistence or rigidity associated with the p factor or level of personality functioning?Evolutionary and developmental research on the origins of social communication converge to suggest that the capacity for epistemic trust, and the capacity for salutogenesis that it enables, may be an important candidate mechanism in this context (Luyten, Campbell, et al., 2020, Luyten, Campbell, & Fonagy, 2020).Epistemic trust refers to the capacity of humans to identify knowledge conveyed by others as personally relevant and generalizable to other contexts.It enables a kind of social learning that is largely species-specific, as it involves encoding knowledge offered by others as significant, relevant to the recipient, and socially generalizable, and thus enables forms of communication and collaboration that are largely or completely absent in other animals-even in the closest relatives of humans in the animal kingdom, the great apes (Tomasello & Vaish, 2013).The body of research underpinning these views, although extensive, has surprisingly been largely left unintegrated with contemporary approaches to personality and psychopathology (Csibra & Gergely, 2009;Sperber et al., 2010;Tomasello, 2010).
Developmental research has shown that epistemic trust emerges only slowly, in the first few years of life.By contrast, epistemic vigilance-that is, lack of trust in knowledge offered by others-quickly emerges as the default position in children (Mascaro & Sperber, 2009;Sperber et al., 2010).Hence, epistemic vigilance needs to be overcome and epistemic trust needs to be established (Wilson & Sperber, 2012).Attachment figures appear to play central role in this process, as they are key in putting the developing child into a so-called "learning mode" by using particular communicative signals, termed ostensive cues, which trigger epistemic openness in the child (Csibra & Gergely, 2009).Hence, within attachment relationships the developing child acquires the capacity to discern the epistemic trustworthiness of others and epistemic trust leads to a virtuous cycle characterized by openness to others and thus, more generally, to positive influences in the environment.
In contrast, there is now good evidence to suggest that early adversity, in terms of either insecure attachment experiences and/or a broader environment marked by insecurity and mistrust, may impair the capacity for epistemic trust, leading to insecure attachment and problems with mentalizing (or reflective functioning), effectively closing off the individual from others and the wider social environment (Luyten, Campbell, et al., 2020, Luyten, Campbell, & Fonagy, 2020).As a result, the individual finds him/herself isolated from others and thus is cut off from opportunities to recalibrate his/her mind in interactions with others.Consistent with longstanding theoretical and empirical traditions emphasizing the role of problems with (social) learning and rigidity in psychopathology reviewed elsewhere (Fonagy et al., 2015), this state of epistemic vigilance/distrust may underlie both the persistence of mental disorders and the assumed "rigidity", "pathological personality traits", "insecure attachment styles", or "mentalizing impairments" that are assumed to trigger psychopathology.Because these features appear to be relatively (and sometimes markedly) stable across time and contexts, the pitfall is that laypersons, scientists, and clinicians alike attribute these characteristics to the individual rather than to features of the relationship of the individual with his/her environment.
However, research findings reviewed above suggest that epistemic vigilance is best seen as an understandable adaptation strategy given the environment that many people grow up in.As an example, children who present with conduct disorder typically show problems with the "unlearning" of aggression (Fonagy & Luyten, 2018).However, for many young people growing up in a social environment marked by abuse, neglect, and violence, high levels of aggression simply represent an appropriate survival strategy, and, from the perspective of these young people, trusting others may actually compromise their survival.Hence, these young people "are" not aggressive, and their high levels of aggression are not necessarily stable across time and contexts (as is demonstrated, for instance, by the fact that many of these young people can be caring, protective, and loving towards some others in their lives), but they use excessive aggression as a coping strategy to deal with experiences of abuse and neglect, and associated feelings of worthlessness and desperation.Hence, what we argue is that the "personality" or "personality disorder" of these individuals is stable only insofar as the mechanisms that underpin their coping strategy is active.
These views have important implications not only for theories concerning the relationship between personality and psychopathology, but also for prevention and treatment strategies.Of course, once again, we are mindful not to fall prey to a naive environmental position.There is good evidence to suggest that genetic predisposition, most likely mediated by temperamental factors such as effortful control and executive functioning (Fonagy & Luyten, 2018) and broader environmental factors-in particular, social inequality-may influence the development of epistemic trust and the capacity for salutogenesis.More research in this area is definitely needed, but findings concerning the massive influence of social inequality on the prevalence, persistence, and outcome of mental health problems, as well as epistemic trust (Rözer et al., 2016;Rözer & Volker, 2016), urge us to reconsider the role of the broader sociocultural environment in the development of both personality and psychopathology.The social environment, in interaction with biological predisposition, is more than likely to nudge individuals in a certain direction in the epigenetic landscape and/or put constraints on the flexibility for change in the course of development.This also explains why there is probably only a fairly limited number of prototypical ways of negotiating developmental tasks, and thus a fairly limited number of personality prototypes, as the number of prototypical ways to deal with these conflicts is constrained by biological and social factors (Westen et al., 2006).

IMPLICATIONS
In this article, we have summarized psychodynamic approaches to personality in relation to psychopathology.Both historical and, more pertinently, contemporary psychodynamic approaches are based on the assumption of a fundamental continuity between normal and disrupted personality development, and between personality and psychopathology.Specifically, personality-and psychopathology, for that matter-is considered to be an attempt at adaptation to a specific environment, and the typically observed relative stability of "personality traits" is thus thought to largely result from the relative stability of the environment.Hence, overall, it seems difficult to separate the "person" from their "disorder".Recent social-communicative approaches to personality development that conceptualize personality and psychopathology as a problem of interaction with the environment move even further away from the view that there is a neat distinction between personality and psychopathology.
However, although the bulk of research in this area supports these assumptions, we have pointed out that true taxons may exist, and thus, in some instances, it may be possible to distinguish the person from the disorder.In fact, the distinction between the person and the disorder may be moderated by severity (i.e., a general psychopathology or "p" factor) such that increasing levels of severity appear to result in increased intrinsic coupling between the two.At low levels of a general psychopathology factor, psychological disorders might not only occur in relative isolation, but also be relatively distinct from premorbid personality traits.At higher levels of severity, it appears to become impossible to distinguish the person from the disorder.Moreover, we cited evidence to suggest that biological factors such as temperament may be an important moderator of person-environment exchanges, which opens up perspectives for the integration of the views expressed in this article and multivariate personality approaches that focus on the hierarchical structure of broad-bandwidth personality traits.Culture might be an important moderator of this relationship in that in some cultures personality features associated with higher levels of p are still considered to reflect normal personality variations (for instance, psychotic behaviors are considered to be less pathological in some cultures compared to others).More research in non-Western cultures is therefore needed, as most research on personality is still conducted in so-called WEIRD (Western, Educated, Industrialized, Rich and Democratic) people, although they constitute only a small proportion of the world population (Henrich, 2020).
Future research should also focus on the integration between different views and findings on the structure of personality and pathology in a more encompassing theory of human development.Moreover, with regard to prevention and intervention, the approach to the development of personality and psychopathology summarized in this article has important clinical implications.First, these views suggest that broad transdiagnostic treatments that address basic personality issues underlying psychopathology may be more effective and productive than treatments that are based on the assumption that psychological disorders each have their own etiology (Caspi et al., 2014;Luyten et al., 2017).Second, a factor common to all effective treatments may be that they address the apparent rigidity associated with psychopathology and help to bring about changes in person-environment transactions, opening up the capacity for salutogenesis (Fonagy et al., 2017b).There may be a fundamental parallel between | 11 LUYTEN aNd FONaGY normal psychological development and the processes of therapeutic change (Blatt, 2008).Different psychosocial interventions may provide patients with opportunities to examine their relation to others and their broader social environment in a context of feeling understood and appreciated by an other, which allows patients to develop more reciprocal and mutually facilitating interpersonal relationships, leading to greater openness and flexibility.However, both biological and environmental factors may constrain this process.Hence, in our opinion, the future of personality-focused approaches to psychopathology lies in investigating the factors that determine the developmental pathways involved in the flexibility and rigidity of the individual in relation to environmental factors.

F
Relationship between the general psychopathology factor, levels of personality functioning, and persistence of psychopathology (Adapted fromCaspi et al., 2014)