Introduction

For many in the public at large, the term “psychopath” conjures up images of ruthless homicidal maniacs and criminal masterminds. This impression is reinforced on an ongoing basis by depictions of psychopathic individuals in popular books and films, such as No Country for Old Men, Silence of the Lambs, and Catch Me if You Can, and by media accounts of high-profile criminals ranging from Charles Manson to Jeffrey Dahmer to Bernie Madoff. However, the concept of psychopathy (“psychopathic personality”) held by experts in the mental health field differs sharply from this common public perception—emphasizing distinct dispositional tendencies as opposed to serious criminal acts of one sort or another. This module reviews historic and contemporary conceptions of psychopathy as a clinical disorder, describes methods for assessing it, and discusses how a new conceptual model can help to address key questions regarding its nature and origins that have long been debated. It will be seen from this review that the topic remains no less fascinating or socially relevant when considered from a clinical–scientific perspective.

Historic Conceptions

Early writers characterized psychopathy as an atypical form of mental illness in which rational faculties appeared normal but everyday behavior and social relationships are markedly disrupted. French physician Philippe Pinel (1806/1962) documented cases of what he called manie sans delire (“insanity without delirium”), in which dramatic episodes of recklessness and aggression occurred in individuals not suffering from obvious clouding of the mind. German psychiatrist Julius Koch (1888) introduced the disease-oriented term psychopathic to convey the idea that conditions of this type had a strong constitutional-heritable basis. In his seminal book The Mask of Sanity, which focused on patients committed for hospital treatment, American psychiatrist Hervey Cleckley (1941/1976) described psychopathy as a deep-rooted emotional pathology concealed by an outward appearance of good mental health. In contrast with other psychiatric patients, psychopathic individuals present as confident, sociable, and well adjusted. However, their underlying disorder reveals itself over time through their actions and attitudes. To facilitate identification of psychopathic individuals in clinical settings, Cleckley provided 16 diagnostic criteria distilled from his clinical case summaries, encompassing indicators of apparent psychological stability (e.g., charm and intelligence, absence of nervousness) along with symptoms of behavioral deviancy (e.g., irresponsibility, failure to plan) and impaired affect and social connectedness (e.g., absence of remorse, deceptiveness, inability to love).

Notably, Cleckley did not characterize psychopathic patients as inherently cruel, violent, or dangerous. Although some engaged in repetitive violent acts, more often the harm they caused was nonphysical and the product of impulsive self-centeredness as opposed to viciousness. Indeed, Cleckley’s case histories included examples of “successful psychopaths” who ascended to careers as professors, medical doctors, or businessmen, along with examples of more aimless dysfunctional types. In contrast with this, other writers from Cleckley’s time who were concerned with criminal expressions of psychopathy placed greater emphasis on symptoms of emotional coldness, aggression, and predatory victimization. For example, McCord and McCord (1964) described the condition in more generally pathologic terms, highlighting “guiltlessness” (lack of remorse) and “lovelessness” (lack of attachment capacity) as central defining features.

Cleckley’s conception served as a referent for the diagnosis of psychopathy in the first two editions of the official American psychiatric nosology, the Diagnostic and Statistical Manual of Mental Disorders (DSM). However, a dramatic shift occurred in the third edition of the DSM, with the introduction of behaviorally oriented symptom definitions for most disorders to address longstanding problems of reliability. The Cleckley-oriented conception of psychopathy in prior editions was replaced by antisocial personality disorder (ASPD), defined by specific indicants of behavioral deviancy in childhood (e.g., fighting, lying, stealing, truancy) continuing into adulthood (manifested as repeated rulebreaking, impulsiveness, irresponsibility, aggressiveness, etc.). Concerns with this new conception were expressed by psychopathy experts, who noted that ASPD provided limited coverage of interpersonal-affective symptoms considered essential to psychopathy (e.g., charm, deceitfulness, selfishness, shallow affect; Hare, 1983). Nonetheless, ASPD was retained in much the same form in the fourth edition of the DSM (DSM-IV; American Psychiatric Association [APA], 2000), and remained unchanged in the fifth edition of the DSM (American Psychiatric Association, 2013). That said, the DSM-5 does include a new, dimensional-trait approach to characterizing personality pathology (Strickland, Drislane, Lucy, Krueger, & Patrick, in press).

Contemporary assessment methods

Modern approaches to the assessment of psychopathy, consisting of rating instruments and self-report scales, reflect the foregoing historic conceptions to differing degrees.

Psychopathy in adult criminal offenders

The most widely used instrument for diagnosing psychopathy in correctional and forensic settings is the Psychopathy Checklist-Revised (PCL-R; Hare, 2003), which comprises 20 items rated on the basis of interview and file-record information. The items of the PCL-R effectively capture the interpersonal-affective deficits and behavioral deviance features identified by Cleckley, but include only limited, indirect coverage of positive adjustment features. The manual for the PCL-R recommends the use of a cutoff score of 30 out of 40 for assigning a diagnosis of psychopathy. High overall PCL-R scores are correlated with impulsive and aggressive tendencies, low empathy, Machiavellianism, lack of social connectedness, and persistent violent offending. Given these correlates, and the omission of positive adjustment indicators, psychopathy as assessed by the PCL-R appears more similar to the predatory-aggressive conception of McCord and McCord than to Cleckley’s conception.

Although the PCL-R was developed to index psychopathy as a unitary condition, structural analyses of its items reveal distinct interpersonal-affective and antisocial deviance subdimensions (factors). Although moderately (about .5) correlated, these factors show contrasting relations with external criterion measures. The interpersonal-affective factor relates to indices of narcissism, low empathy, and proactive aggression (Hare, 2003), and to some extent (after controlling for its overlap with the antisocial factor) adaptive tendencies such as high social assertiveness and low fear, distress, and depression (Hicks & Patrick, 2006). High scores on the antisocial deviance factor, by contrast, are associated mainly with maladaptive tendencies and behaviors, including impulsiveness, sensation seeking, alienation and mistrust, reactive aggression, early and persistent antisocial deviance, and substance-related problems.

Psychopathy in noncriminal adults

Psychopathy has most typically been assessed in noncriminal adult samples using self-report-based measures. Older measures of this type emphasized the antisocial deviancy component of psychopathy with limited coverage of interpersonal-affective features (Hare, 2003). Some newer instruments provide more balanced coverage of both. One example is the now widely used Psychopathic Personality Inventory (PPI; Lilienfeld & Andrews, 1996), which was developed to index personality dispositions embodied within historic conceptions of psychopathy. Its current revised form (PPI-R; Lilienfeld & Widows, 2005) contains 154 items, organized into eight facet scales.

Like the items of the PCL-R, the subscales of the PPI cohere around two distinguishable factors: a fearless dominance (FD) factor reflecting social potency, stress immunity, and fearlessness, and a self-centered impulsivity (SCI) factor reflecting egocentricity, exploitativeness, hostile rebelliousness, and lack of planning. However, unlike the factors of the PCL-R, the two PPI factors are uncorrelated, and thus even more distinct in their external correlates. Scores on PPI-FD are associated with indices of positive psychological adjustment (e.g., higher well-being; lower anxiety and depression) and measures of narcissism (low) empathy, and thrill/adventure seeking (Benning, Patrick, Blonigen, Hicks, & Iacono, 2005). Given this, PPI-FD has been interpreted as capturing a more adaptive expression of dispositional fearlessness (i.e., boldness; see below) than the interpersonal-affective factor of the PCL-R—which can be viewed as tapping a more pathologic (antagonistic or “mean”) expression of fearlessness. Scores on PPI-SCI, like Factor 2 of the PCL-R, are associated with multiple indicators of deviancy—including impulsivity and aggressiveness, child and adult antisocial behavior, substance abuse problems, heightened distress and dysphoria, and suicidal ideation.

Psychopathy in child and adolescent clinical samples

Different inventories exist for assessing psychopathic tendencies in children and adolescents. The best-known consist of rating-based measures developed, using the PCL-R as a referent, to identify psychopathic individuals among youth convicted of crimes or referred for treatment of conduct problems. The emphasis in work of this type has been on the importance of psychopathic features for predicting greater severity and persistence of conduct problems. Termed “callous-unemotional” traits, these features encompass low empathy, deficient remorse or guilt, shallow affect, and lack of concern about performance in school and other contexts (Frick & Moffitt, 2010).

One extensively researched measure for assessing psychopathic tendencies in youth is the Antisocial Process Screening Device (APSD; Frick & Marsee, 2006), used with clinic-referred children ages 6 through 13. The APSD includes 20 items completed by parents or teachers. As with the PCL-R and PPI, the items of the APSD tap two distinct factors: a Callous-Unemotional (CU) traits factor, reflecting emotional insensitivity and disregard for others; and an Impulsive/Conduct Problems (I/CP) factor, reflecting impulsivity, behavioral deviancy, and inflated self-importance. Children high on the I/CP factor alone show below-average intelligence, heightened emotional responsiveness to stressors, and angry (reactive) aggression (Frick & Marsee, 2006). By contrast, children high on both APSD factors show average or above-average intelligence, low reported levels of anxiety and nervousness, reduced reactivity to stressful events, and preference for activities entailing novelty and risk. They also learn less readily from punishment and engage in high levels of premeditated as well as reactive aggression and exhibit more persistent violent behavior across time. Given the documented importance of CU traits in moderating the expression of conduct disorder, the upcoming fifth edition of the DSM will include criteria for designating a distinct CU variant of child conduct disorder (Frick & Moffitt, 2010).

Core ingredients of psychopathy: disinhibition, boldness, and meanness

The foregoing material highlights the fact that historic conceptions of psychopathy and available instruments for assessing it place differing emphasis on different symptomatic features. This had contributed to longstanding disagreements among scholars about what psychopathy entails and what causes it. A theoretic conceptualization formulated recently to reconcile alternative perspectives is the Triarchic model (Patrick, Fowles, & Krueger, 2009). This model conceives of psychopathy as encompassing three separable symptomatic components—disinhibition, boldness, and meanness—that can be viewed as thematic building blocks for differing conceptions of psychopathy.

Definitions

Disinhibition as described in the Triarchic model encompasses tendencies toward impulsiveness, weak behavioral restraint, hostility and mistrust, and difficulties in regulating emotion. Meanness entails deficient empathy, lack of affiliative capacity, contempt toward others, predatory exploitativeness, and empowerment through cruelty and destructiveness. Referents for disinhibition and meanness include the finding of distinct I/CP and CU factors in the child psychopathy literature and corresponding evidence for distinct disinhibitory and callous-aggression factors underlying impulse control (externalizing) problems in adults (Krueger, Markon, Patrick, Benning, & Kramer, 2007). The third construct in the model, Boldness,encompasses dominance, social assurance, emotional resiliency, and venturesomeness. Referents for this construct include the “mask” elements of Cleckley’s conception, Lykken’s (1995) low fear theory of psychopathy, the FD factor of the PPI, and developmental research on fearless temperament as a possible precursor to psychopathy (Patrick et al., 2009).

From the perspective of the Triarchic model, Cleckley’s conception of psychopathy emphasized boldness and disinhibition, whereas criminally oriented conceptions (and affiliated measures, including the PCL-R and APSD) emphasize meanness and disinhibition more so. According to the model, individuals high in disinhibitory tendencies would warrant a diagnosis of psychopathy if also high in boldness or meanness (or both), but individuals high on only one of these tendencies would not. Individuals with differing relative elevations on these three symptomatic components would account for contrasting variants (subtypes) of psychopathy as described in the literature (Hicks, Markon, Patrick, Krueger, & Newman, 2004; Karpman, 1941; Skeem, Johansson, Andershed, Kerr, & Louden, 2007).

An inventory designed specifically to operationalize this model is the Triarchic Psychopathy Measure (TriPM; Patrick, 2010). The TriPM contains 58 items comprising three subscales that correspond to the constructs of the model (see Table 1). The items of the Disinhibition and Meanness scales (20 and 19 items, respectively) are taken from the Externalizing Spectrum Inventory (ESI; Krueger et al., 2007), a measure of problems and traits associated with externalizing psychopathology. The TriPM Boldness scale was developed to index fearless tendencies in social, affective-experiential, and activity preference domains, with reference to the FD factor of the PPI and the general factor shown to underlie differing scale measures of fear and fearlessness (Kramer, Patrick, Gasperi, & Krueger, 2012).

Although the TriPM is relatively new, promising evidence for its convergent and discriminant validity has begun to appear (e.g., Sellbom & Phillips, 2013; Strickland et al., in press; see also Venables & Patrick, 2012). Given that the inventory is freely available online, and that several foreign-language translations now exist (including Brazilian-Portuguese, Dutch, Finnish, German, Italian, Portuguese, Swedish, and Spanish), it can be expected that additional validity data will accumulate rapidly over time. Work is also being done to evaluate whether effective scale measures of the Triarchic constructs can be derived from items of other existing psychopathy inventories such as the PPI. As discussed in the last part of this section, research examining the common and distinctive correlates of these three components of psychopathy is likely to be helpful for addressing and perhaps resolving ongoing points of uncertainty and debate in the field.

Causal factors

Considerable research has been devoted over many years to investigation of causal factors in psychopathy. Existing theories are of two types: (1) theories emphasizing core deficits in emotional sensitivity or responsiveness, and (2) theories positing basic impairments in cognitive-attentional processing (Patrick & Bernat, 2009). In support of these alternative theories, differing neurobiological correlates of psychopathy have been reported. One of the most consistent entails a lack of normal enhancement of the startle blink reflex to abrupt noises occurring during viewing of aversive foreground stimuli (e.g., scary or disturbing pictorial images) as compared with neutral or pleasant stimuli (see Figure 1). This result, akin to a failure to “jump” upon hearing a trash can tip while walking alone in a dark alley, has been interpreted as reflecting a lack of normal defensive (fear) reactivity. Another fairly consistent finding involves reduced amplitude of brain potential response to intermittent target stimuli, or following incorrect responses, within cognitive performance tasks—indicative of reduced cortical-attentional processing or impaired action monitoring (Patrick & Bernat, 2009). Yet other research using functional neuroimaging has demonstrated deficits in basic subcortical (amygdala) reactivity to interpersonal distress cues (e.g., fearful human faces) in high-psychopathic individuals (Jones, Laurens, Herba, Barker, & Viding, 2009; Marsh et al., 2008).

The Triarchic model may prove to be of use for reconciling alternative causal models of psychopathy that have been proposed based on contrasting neurobiological and behavioral findings. For example, lack of startle enhancement during aversive cuing has been tied specifically to the interpersonal-affective factor of the PCL-R and the counterpart FD factor of the PPI (Figure 1)—suggesting a link to the boldness component of psychopathy. By contrast, reduced brain potential responses in cognitive tasks appear more related to impulsive-externalizing tendencies associated with the disinhibition component of psychopathy (Carlson, Thái, & McLaron, 2009; Patrick & Bernat, 2009). On the other hand, the finding of reduced subcortical response to affective facial cues has been tied to the CU traits factor of child/adolescent psychopathy, a referent for meanness in the Triarchic model. However, further research is needed to determine whether this finding reflects fear deficits common to meanness and boldness, or deficits in affiliative capacity or empathy specific to meanness.

Triarchic model perspective on long-debated issues regarding psychopathy

As highlighted in the foregoing sections, scholars have grappled with issues of definition since psychopathy was first identified as a condition of clinical concern, and questions regarding its essential features and alternative expressions continue to be debated and studied. This final subsection discusses how some of the major issues of debate are addressed by the Triarchic model.

One key issue is whether psychological/emotional stability is characteristic or not of psychopathy. Cleckley’s (1941/1976) view was that psychopathy entails a salient presentation of good mental health, and his diagnostic criteria included indicators of positive adjustment. By contrast, the dominant clinical assessment devices for psychopathy, the PCL-R and ASPD, are heavily oriented toward deviancy and include no items that are purely indicative of adjustment. From a Triarchic model standpoint, the more adaptive elements of psychopathy are embodied in its boldness facet, which entails social poise, emotional stability, and enjoyment of novelty and adventure. At the same time, high boldness is also associated with narcissistic tendencies, reduced sensitivity to the feelings of others, and risk-taking (Benning et al., 2005). Thus, the concept of boldness provides a way to think about the intriguing “mask” element of psychopathy.

Related to this, another issue is whether lack of anxiety is central to psychopathy, as Cleckley and others (e.g., Fowles & Dindo, 2009; Lykken, 1995) have emphasized. This perspective is challenged by research showing either negligible or somewhat positive associations for overall scores on the PCL-R and other psychopathy measures with anxiety. The Triarchic model helps to address this inconsistency by separating the disorder into subcomponents or facets, which relate differently to measures of trait anxiety: Boldness is correlated negatively with anxiousness (Benning et al., 2005), whereas Disinhibition and Meanness are correlated negatively and negligibly, respectively, with anxiety (Venables & Patrick, 2012). Related to this, cluster analytic studies of criminal offenders exhibiting high overall scores on the PCL-R have demonstrated one subtype characterized by low anxiety in particular, and another exhibiting high anxiety along with very high levels of impulsivity and aggression (Hicks et al., 2004; Skeem et al., 2007). The implication is that low anxiousness is central to one variant of criminal psychopathy (the bold-disinhibited, or “primary” type) but not to another variant (the “disinhibited-mean,” “aggressive-externalizing,” or “secondary” type).

A further key question is whether violent/aggressive tendencies are typical of psychopathic individuals and should be included in the definition of the disorder. Cleckley’s (1941/1976) view was that “such tendencies should be regarded as the exception rather than as the rule” (p. 262). However, aggressiveness is central to criminally oriented conceptions of psychopathy, and the PCL-R includes an item reflecting hot-temperedness and aggression (“poor behavioral controls”) along with other items scored in part based on indications of cruelty and violence. In the Triarchic model, tendencies toward aggression are represented in both the disinhibition and meanness constructs, and a “mean-disinhibited” type of psychopath clearly exists, marked by the presence of salient aggressive behavior (Frick & Marsee, 2006; Hicks et al., 2004). Thus, Cleckley’s idea of aggression as ancillary to psychopathy may apply more to a variant of psychopathy that entails high boldness in conjunction with high disinhibition (Hicks et al., 2004).

Another question is whether criminal or antisocial behavior more broadly represents a defining feature of psychopathy, or a secondary manifestation (Cooke, Michie, Hart, & Clark, 2004). From the standpoint of the Triarchic model, antisocial behavior arises from the complex interplay of different “deviance-promoting” influences—including dispositional boldness, meanness, and disinhibition. However, whether approaches can be developed for classifying antisocial behaviors in ways that relate more selectively to these and other distinct influences (e.g., through reference to underlying motives, spontaneity versus premeditation) is an important topic to be addressed in future research.

Another key question is whether differing subtypes of psychopathy exist. From the perspective of the Triarchic model, alternative variants of psychopathy reflect differing configurations of boldness, meanness, and disinhibition. Viewed this way, designations such as “bold-disinhibited” and “mean-disinhibited” may prove more useful for research and clinical purposes than labels like “primary” versus “secondary” or “low anxious” versus “high anxious.” An issue from this perspective is whether individuals who are high in boldness and/or meanness but low in disinhibition would qualify for a diagnosis of psychopathy. For example, should a high-bold/high-mean individual (e.g., a ruthless corporate executive, like the one portrayed by actor Michael Douglas in the film Wall Street; Pressman & Stone, 1987)—or an extremely mean/vicious but neither bold nor pervasively disinhibited individual, such as Russian serial murder Andrei Chikatilo (Cullen, 1993)—be considered psychopathic? Questions of this sort will need to be addressed through elaborations of existing theories in conjunction with further systematic research.

Yet another question is whether psychopathy differs in women as compared to men. Cleckley’s descriptive accounts of psychopathic patients included two female case examples along with multiple male cases, and his view was that psychopathy clearly exists in women and reflects the same core deficit (i.e., absence of “major emotional accompaniments” of experience) as in men. However, men exhibit criminal deviance and ASPD at much higher rates than women (APA, 2000) and men in the population at large score higher in general on measures of psychopathy than women (Hare, 2003; Lilienfeld & Widows, 2005). From a Triarchic model perspective, these differences in prevalence may be attributable largely to differences between women and men in average levels of boldness, meanness, and disinhibition. Some supportive evidence exists for this hypothesis (e.g., findings of Hicks et al. [2007] demonstrating mediation of gender differences in ASPD symptoms by levels of externalizing proneness). Beyond this, it is important also to consider whether underlying psychopathic dispositions in men and women may be manifested differently in overt behavior (Verona & Vitale, 2006). Some intriguing evidence exists for this—including twin research findings demonstrating a genetic association between dispositional boldness (as indexed by estimated scores on PPI-FD) and a composite index of externalizing problems in male but not female participants (Blonigen, Hicks, Patrick, Krueger, Iacono, & McGue, 2005). However, more extensive research along these lines, examining all facets of the Triarchic model in relation to behavioral outcomes of differing kinds, will be required to effectively address the question of gender-moderated expression.

A final intriguing question is whether “successful” psychopaths exist. Hall and Benning (2006) hypothesized that successful psychopathy entails a preponderance of certain causal influences (resulting in particular symptomatic features) over others. Drawing on known correlates of PPI-FD (e.g., Benning et al., 2005; Ross, Benning, Patrick, Thompson, & Thurston, 2009) and theories positing separate etiologic mechanisms for differing features of psychopathy (Fowles & Dindo, 2009; Patrick & Bernat, 2009), these authors proposed that the presence of dispositional fearlessness (boldness) may be conducive to success when not accompanied by high externalizing proneness (disinhibition). For example, high-bold/low-disinhibited individuals could be expected to achieve higher success in occupations calling for leadership and/or courage because their psychopathic tendencies are manifested mainly in terms of social effectiveness, affective resilience, and venturesomeness.

Data relevant to this idea come from an intriguing study by Lilienfeld, Waldman, Landfield, Rubenzer, and Faschingbauer (2012), who used personality trait ratings of former U.S. presidents provided by expert historians to estimate scores on the FD and SCI factors of the PPI (Ross et al., 2009). They found that higher estimated levels of PPI-FD (boldness) predicted higher ratings of presidential performance, persuasiveness, leadership, and crisis management ability, whereas higher estimated levels of SCI predicted adverse outcomes such as documented abuses of power and impeachment proceedings. Further research on outcomes associated with high levels of boldness and/or meanness in the absence of high disinhibition should yield valuable new insights into dispositional factors underlying psychopathy and alternative ways psychopathic tendencies can be expressed.