Dispositional Contempt: A First Look at the Contemptuous Person

Abstract

Contempt is a powerful emotion. Marriages fail (Gottman, 1994), coworkers are shamed (Melwani & Barsade, 2011), terrorism is tended toward (Tausch et al., 2011). Despite its importance, contempt has not been investigated at the level of personality. The present research examines how our contemptuous reactions can be conceptualized and measured as a stable individual-difference variable with a range of theoretically predicted correlates. First, we introduce a measure of dispositional contempt, the tendency to look down on, distance, and derogate others who violate our standards. We then unpack the dynamics of dispositional contempt. Across six studies using self-report and emotion elicitation in student and MTurk samples (Ns = 165 to 1,368), we examined its (1) nomological network, (2) personality and behavioral correlates, and (3) implications for relationship functioning. Dispositional contempt was distinguished from tendencies toward related emotions and was most associated with dispositional envy, anger, and hubristic pride. Somewhat paradoxically, dispositional contempt was related to being cold and “superior,” with associations found with narcissism, other-oriented perfectionism, and various antisocial tendencies (e.g., Disagreeableness, Machiavellianism, racism), but was also related to being self-deprecating and emotionally fragile, with associations found with low self-esteem, insecure attachment, and feeling that others impose perfectionistic standards on oneself. Dispositional contempt predicted contemptuous reactions to eliciting film clips, particularly when targets showed low competence/power. Finally, perceiving one’s romantic partner as dispositionally contemptuous was associated with lower commitment and satisfaction. Taken together, results give a first look at the contemptuous person and provide a new organizing framework for understanding contempt.

“There is not in human nature a more odious disposition than a proneness to contempt.”

(Fielding, 1840, p. 712)

A husband meets his wife’s pleas with nonchalant eye rolls. A teenager blurts “Whatever!” People gossip about or avoid others at lunch or parties. Those who disapproved of Margaret Thatcher in life turn their backs on her coffin as it passes. And the scientist who published high-impact papers using fake data had fellow researchers turn their backs on him, too. Ubiquitous in everyday and not-so-everyday life are people’s experiences with contempt, an emotional reaction that is elicited when a person or group violates one’s standards and one looks down on them with the tendency to distance and/or derogate them. The consequences for recipients of contempt are large, ranging from the physiological stress response associated with shame and rejection (Blackhart, Eckel, & Tice, 2007; Dickerson, Gruenewald, & Kemeny, 2004) to subjective distress and drops in self-esteem (Melwani & Barsade, 2011), from getting outcast by group members (Sommer, Williams, Ciarocco, & Baumeister, 2001) to, at the intergroup level, possibly getting bombed (Tausch et al., 2011). Although the consequences for contemners – those who feel and express contempt – are less well-known, research suggests that they are seen as high-status (Keltner, Young, Heerey, Oemig, & Monarch, 1998; Melwani & Barsade, 2013) and may be prone to cardiovascular disease (Rosenberg, Ekman, & Blumenthal, 1998). In marital contexts, both contemners and the contemned alike lose, as contempt is the biggest predictor of divorce (Gottman, 1993, 1994).

Despite the prevalence and importance of contempt, research on this emotion lags behind that on other emotions and is somewhat disparate. At the time of this writing, using “contempt” as a keyword on psycinfo generated only 162 records, with most of these publications in journals of marketing, communication, and other types of applied psychology. In the social-personality literature, controversy persists regarding what typically elicits contempt (e.g., see Rozin, Lowery, Imada, & Haidt, 1999, vs. Hutcherson & Gross, 2011), the word itself is poorly understood (Wagner, 2000), and its non-verbal expression is unreliably recognized (Rosenberg & Ekman, 1995). These factors may frustrate contempt as an object of research. We sought to integrate and advance our understanding of contempt by examining it at the level of personality. We first introduce a measure of dispositional contempt – the Dispositional Contempt Scale (DCS) – highlighting how steps in its development inform theoretical models of contempt. We then unpack the dynamics of dispositional contempt in terms of its correlates and consequences, including in romantic relationship contexts. We conclude by outlining the implications of contempt as a disposition for contempt as an emotion and suggest future directions for researching either.

What is Contempt?

The word “contempt” and its synonyms, “disdain” and “scorn,” smack of the elevated diction found in Victorian novels. Yet, it is an emotional reaction that is familiar in most, if not all, cultures (Ekman & Friesen, 1986; Matsumoto, 1992). In his landmark research on marital relationships, Gottman (1994) explained that contempt involves the belief that one’s spouse is absurd, incompetent, or beneath dignity, and that it is subjectively similar to exasperation, detachment, and cold hatred. Wagner (2000) described contempt as the feeling that a person is beneath consideration and understood it to have at least three components: it is interpersonal, it involves viewing another person negatively, and it involves feeling superior to that person. In seeking to develop an understanding of contempt as a disposition by considering what distinguishes it as an emotion – its antecedent appraisals, action tendencies, and social functions – we built on and extended Wagner’s (2000) definition, elaborating on the content of these negative appraisals and speaking not only to perceiving another’s inferiority but to the rejecting and distancing behaviors that follow. We briefly review these components of contempt as an emotional reaction below.

First, contempt is grounded in detecting others’ violating standards that are important to the contemner and likely to his or her social group. There is division in the literature regarding which domains of functioning these standards concern. This is important to resolve for building an account of dispositional contempt inasmuch as it speaks to what contemptuous people are contemptuous about. In their CAD Triad Hypothesis, Rozin et al. (1997) categorized contempt, along with anger and disgust, as one of the “other-critical moral emotions,” thus viewing contempt as inherently associated with morality. Their work suggested that people react with contempt to violations of community (i.e., respect for hierarchical and communal obligations), with anger to violations of autonomy (i.e., one’s freedom and rights), and with disgust to violations of divinity (i.e., what is sacred and/or pure). Accordingly, we considered content related to immorality, including violations of communal ethics (e.g., “I tend to snub people who have behaved unethically”), in developing the DCS.¹

In tackling the question of how the “moral” emotions of contempt, anger, and disgust are distinguished, however, Hutcherson and Gross (2011) demonstrated a link between contempt and judging incompetence. They found that contempt was evoked by appraising others’ abilities to execute their aims (competence) rather than whether these aims were good or bad (morality), the latter of which affected levels of moral disgust. Similarly, Ufkes et al. (2011) found that, given a vignette of conflict with a hypothetical outgroup, participants’ ratings of contempt for the outgroup were impacted by competence-based stereotype content, whereas warmth-related content affected anger ratings. Following this research, we also included competence-related content in developing our measure (e.g., “When I think others are incompetent, I tend to keep my distance.”). Incorporating both views allowed us to ask if dispositional contempt has a unidimensional structure such that individuals who tend to feel contempt for incompetence also tend to feel it for immorality, or if contemptuous people tend to be concerned with one domain or another.

A second feature of contempt that also relates to its antecedent appraisals is that the offending party is looked down upon. This ties contempt to perceiving another’s inferiority. This also differentiates contempt from anger as a person can still respect someone who has angered them. With contemptuous reactions’ signaling these relative value judgments (Ben-Ze’ev, 2000), contempt is intrinsically linked with status and hierarchy. Research has found that displaying contempt triggers perceptions of strength and superiority over others, causing an imbalance that leads to a boost in status for those expressing contempt and a drop in status for those receiving it (Keltner & Haidt, 1999; Keltner, Young, Heerey, Oemig, & Monarch, 1998). From the standpoint of signaling low value, it is also fitting that the facial expression of contempt (e.g., unilateral lip curl) is often mislabeled as “boredom” (Russell, 1991).

A third feature of contempt as an emotion is its association with avoidant/dismissive action tendencies. Contempt motivates behavior that distances or rejects the contempt target, such as by removing him or her from one’s social group through social ostracism (Fischer & Roseman, 2007) or compromising his or her status in the group through gossip (Galen & Underwood, 1997; Wert & Salovey, 2004). Fischer and Roseman (2007) found that, unlike anger, which is characterized by short-term attacks in the hopes of long-term reparation, contempt involves social rejection and exclusion of a target in the short- as well as long-term. Their research suggested this is because contempt involves developing a more negative view of the target that extends to his or her character such that he or she is deemed incapable of meeting important standards and the contemptuous agent feels unable to change him or her. These distancing tendencies are also found at the intergroup level (Mackie, Devos, & Smith, 2010).

Ultimately, in keeping with social functionalist accounts of emotion (Frijda & Mesquita, 1994; Keltner & Haidt, 1999), contempt serves the distinct social regulatory role of minimizing the negative impact of others by weeding them out or shaping them through threats or cuts to their inclusionary status by distancing and basically shaming them (Descioli & Kurzban, 2009). Indeed, contempt has been noted to be issued in an “attempt, internally or overtly, to reduce the other or others to feeling like nothing, nobody, someone worthless and unacceptable” (Mindell, 1994). The extent to which contempt might entail such a psychological or physical reduction of a contempt target is exemplified by work by Tausch et al. (2011) that demonstrated that having contempt, rather than anger, for a party viewed as threatening or responsible for injustice is associated with the desire to take radical action against them, such as through acts of vandalism, violence, or terror, rather than using normative means such as signing petitions or protesting. For such reasons, being around contemptuous people may have especially dangerous repercussions.

What is Dispositional Contempt?

The central hypothesis of the current research is that people vary in their level of dispositional contempt – that is, in their tendency to look down on, feel cold toward, and derogate or distance those who violate their standards and values. In thinking about contempt as a disposition, it is worth briefly considering affect as a disposition, as a number of emotion dispositions have been investigated. Among these are dispositional envy (Smith et al., 1999), guilt-proneness (Cohen et al., 2011; Tangney & Dearing, 2002), shame proneness (Tangney & Dearing, 2002), trait pride (Tracy & Robins, 2007), dispositional awe-proneness (Shiota, Keltner, & John, 2006), trait anger (Spielberger, 1988), disgust sensitivity (Haidt, McCauly, & Rozin, 1994), nostalgia proneness (Seehusen et al., 2013), spitefulness (Marcus, Zeigler-Hill, Mercer, & Norris, 2014), and dispositional greed (Krekels & Pandelaere, 2015; Seuntjens, Zeelenberg, van de Ven, & Breugelmans, 2015). Unlike emotions or affective states, which are “acute, intense, and typically brief psychophysiological changes that result from a response to a meaningful situation in one’s environment” (Rosenberg, 1998, p. 250), emotion dispositions, or affective traits, are “stable predispositions toward certain types of emotional responding” that “set the threshold for the occurrence of particular emotional states” (p. 249). With emotions as guides, affective biases shape how individuals routinely feel, construe life events, and interact with their environments.

We have several reasons for suspecting that contempt manifests at the level of personality. For one, longitudinal observations of interpersonal interactions have found that contemptuous behaviors – eye-rolling, sarcasm, and so on – show remarkable rank-order stability across individuals. The stability of contemptuous behaviors observed in couples whose visits to the laboratory were separated by four years was high for husbands (stability coefficient = .76) and moderately so for wives (.46) (Gottman & Levenson, 1999), suggesting that the degree to which people experience and express contempt, at least in marital contexts, is consistent. Second, given that contempt stems from social appraisals and that people are relatively stable in (1) their social-cognitive and affective processing biases (Mischel & Shoda, 1995) and (2) the standards and values they endorse (Schwartz, 1992), the emergence of contempt as a trait-like variable may be expected. Third, compared to other emotions, contempt tends to linger (Fischer & Roseman, 2007). This may make contempt more likely to mold one’s thoughts, feelings, and behaviors toward others and manifest at the level of stable individual differences. Finally, the social situations that elicit contempt offer a wide range of alternate responses – compassion, pity, forgiveness, tolerance – against which biases may emerge that gain stability over time as individuals grow accustomed to responding in one way or another. Individuals low in the motivation and ability to get along with others, for example, may turn to contempt, evolutionarily designed to save resources, as a cost-effective coping mechanism.

How Does Dispositional Contempt Relate to Other Dispositions?

Assuming that individual differences in dispositional contempt exist, what are the contemptuous like? What characteristics define contemptuous people such that we can learn something about how they operate in the social milieu and possibly what fuels their contempt? The social-personality literature already contains a host of theoretically related constructs, such as hubristic pride (Tracy & Robins, 2004), Disagreeableness (Costa & McCrae, 1992), and narcissism (Morf & Rhodewalt, 2001). By specifying linkages between dispositional contempt and these constructs, we can use what has been established about them to gain a better understanding of contempt-proneness as a chronically maintained process. At the same time, because associations of dispositional contempt with similar tendencies, if strong, would cast doubt on whether dispositional contempt is more than a relabeling of them, a first step after developing the DCS was to establish its divergent validity.

We now present a theoretical account of dispositional contempt that integrates what has been established about contempt as an emotion to predict what defines it as a disposition. In so doing, we offer specific hypotheses regarding (1) interrelated emotion dispositions and personality dimensions, (2) narrower attributes that define contemptuous people based on social-cognitive and motivational tendencies likely to promote contempt, and (3) intra- and interpersonal consequences of contempt-proneness, including in romantic relationships. We viewed each area as foundational to understanding contempt as a disposition and address each in our empirical work.

Other Emotion Dispositions

The subjective and appraisal elements of contempt (e.g., unpleasant; social aversion; inflated self) render contempt similar to other emotions. Contempt has elements of “shame–‘the other should not be doing what she is doing’; anger–‘the other is doing something against me’; irritation–‘the other is doing something to make it inconvenient for me’; snobbery–‘the other is simply not as good as I am” (Fredericson, 2010, p. 233). Indeed, contempt has often been considered a mixture or variant of anger and disgust due to research showing that its facial expression is confused with anger and disgust (Frijda & Tcherkassof, 1997; Russell, 1991; Shioiri et al., 1999) and that its experience is similar to or co-occurs with these emotions (Fischer & Roseman, 2007). This makes contempt conceptually interesting but presents a challenge to establishing contempt-proneness as something incrementally useful for explaining and predicting behavior. Thus, we investigated the nomological network of dispositional contempt with regard to other emotion dispositions and their differential relations with other personality traits.

Dispositional Contempt and Broader Personality Dimensions

Personality traits themselves are seen as enduring patterns of feelings, thoughts, and behaviors that reflect characteristic ways of interacting with the world (Roberts & Jackson, 2008; Tellegen, 1991). Individual differences in emotions, like contempt, are thus deemed key ingredients of personality that help organize its stable aspects (Izard, 1977; Magai & Haviland-Jones, 2002; Magai & Nusbaum, 1996; Tomkins, 1962, 1963). We investigated the association of contempt-proneness with major dimensions of personality in hoping to differentiate dispositional contempt from similar emotion dispositions but also in moving toward the goal of understanding it as a stable configuration of feelings, thoughts, and behaviors that likely relates to many areas of personality functioning. Although contempt is a normal and adaptive reaction that serves to regulate social standards, dispositional contempt implies being at the extreme end of a continuum. How does that extremity manifest in personality?

Invoking the Big Five or Five-Factor Model taxonomy (John, Naumann, & Soto, 2008; McCrae & Costa, 2008), we expected a robust relation between contempt-proneness and low Agreeableness. Agreeableness involves the motivation to maintain interpersonal positive relations and is manifested in tendencies toward warmth, empathy, altruism, generosity, cooperation, and politeneness (Graziano & Tobin, 2009). Most descriptions of Agreeableness do not reveal overt links to contempt, but behaviors and characteristics at its negative pole – being fault-finding and insulting, cold and unsympathetic – are relevant to individuals who would sooner distance or derogate others for their perceived flaws rather than try to get along with them, such as by helping or tolerating them. Disagreeable people experience high conflict and poor social functioning (Jensen-Campbell & Graziano, 2001), and it may be that contempt, toxic to relationships, is a key mechanism for understanding these outcomes. Other Big Five traits have less obvious conceptual links to contempt, but we expected contemptuous people to be more neurotic given the negative reactivity involved in contemptuous responding.

The disagreeable nature of contempt-proneness can be further unpacked by examining relations between contempt-proneness and the “Dark Tetrad” (Buckels, Jones, & Paulhus, 2013; Chabrol et al., 2009; Furnham et al., 2013). These antisocial traits have callous exploitation at their core (Jakobwitz & Egan, 2006; Jones & Figueredo, 2013) but are conceptually distinct and predict different behaviors (Furnham et al., 2013). While contempt figures prominently in accounts of antisociality (e.g., Bursten, 1972; Millon, Simonsen, Birket-Smith, & Davis, 2003; Tromanhauser, 1989), it has not been systematically studied in that context and may be differentially associated with having inflated self-views (narcissism), pursuing power at others’ expense (Machiavellianism), having cold disregard for others (psychopathy), or feeling pleasure at others’ pain (sadism). We also considered links with covert narcissism, which, like overt narcissism, features excessive self-absorption but is more clearly rooted in insecurity and self-doubt (Wink, 1991). Indeed, the person who looks down on others may not necessarily look up to him or herself. Thus, we viewed the areas of self- vs. other-evaluation as potentially informative regarding the personality dynamics of contempt-proneness.

Social Evaluative Tendencies That Facilitate Dispositional Contempt

The appraisal content and social functions of contempt suggest a number of evaluative biases that define contempt-prone people, and we focused on contributions from perfectionism and self-esteem. If contempt is elicited by appraising others as having violated one’s standards, then characteristically having standards that are excessively high, rigid, and/or vigilantly monitored – the case of perfectionism – should confer vulnerability to contempt. Horney (1950) viewed the perfectionist as “possessing an arrogant contempt for others and as neurotically using high standards as a basis for looking down on others” (Slaney, Ashby, & Trippi, 1995, p. 280). Hewitt and Flett (1991) proposed three variants of perfectionism: self-oriented perfectionism (having high standards for oneself); other-oriented perfectionism (having high standards for others); and socially-prescribed perfectionism (feeling others impose high standards on oneself). Expecting others to be perfect and criticizing them for their missteps ties other-oriented perfectionism to dispositional contempt and should further distinguish it from related constructs.

Regarding self-esteem, on the one hand, contemptuous people may have high self-esteem to the extent that viewing others as deficient along an important parameter elevates their own self-views. Also, high competencies and talents that promote self-esteem may be what give contemptuous people room for looking down on others. On the other hand, contempt-prone individuals may have low self-esteem if their standards are so high that even they cannot reach them, if they are generally harsh judges, or if a history of having been dealt contempt, which lowers self-esteem and breeds contempt (Melwani & Barsade, 2011), is what spurred their own contemptuousness. Ultimately, research on defensive self-esteem regulation (Campbell & Sedikides, 1999) gave us a basis for considering dispositional contempt part of a motivated process associated with low self-esteem. Indeed, it is low-, not high-, self-esteem individuals who tend to profit from making downward social comparisons (Gibbons & Gerrard, 1989). Moreover, perceiving the self as powerless may promote resorting to demeaning and disengaging from others as a method of dealing with them. In light of the interpersonal nature of contempt, we finally turn to the tenuous relationships likely encountered by the contempt-prone.

Disrupted Social Relations and Dispositional Contempt

Contemptuous reactions are intrinsically social and relational, and the last area of social functioning to which our account of dispositional contempt extends is relationships. Here, the overarching idea is that dispositional contempt is related to social disruption. This disruption can be characterized along the interpersonal axes of social affiliation versus power (Bakan, 1966; Wiggins, 1979). The question of contempt-proneness along the vertical dimension of power, initiated by looking at links with Machiavellianism, can be extended to the intergroup level, as tending to find others as inferior may extend to viewing entire groups as inferior. Izard (1977) noted that contempt may target outgroups and be a primary emotional pathway for prejudice. Prejudice itself has been described as “a uniform antipathy or contempt toward an out-group” (Fiske et al., 2002, p. 878, emphasis added). Accordingly, we explored the association of dispositional contempt with racism as well as social dominance orientation (Pratto, Sidanius, Stallworth, & Malle, 1994), which, like racism, reflects the belief that some groups are inherently inferior to others and that hierarchical relations should be maintained.

To better understand how contempt-proneness relates to getting close to others along the horizontal axis of affiliation, we turned to attachment theory, according to which individuals are securely or insecurely attached based on their levels of avoidance (of intimacy and connection) and anxiety (over being rejected and abandoned) (Shaver & Mikulincer, 2004). Attachment avoidance has already been associated with contemptuous behaviors in both sexes (Magai et al., 1995) and with affect “minimization” (Cassidy, 1994) and suppression (Fraley & Shaver, 1997), consistent with the distancing role of contempt. We also considered a relation with attachment anxiety inasmuch as the contempt-prone may be hypervigilant against the same kind of rejection they grant others and inasmuch as their high expectations may extend to high relational needs. Interestingly, in intergroup contexts, it is attachment anxiety, not avoidance, that has been related to hostile and avoidant reactions to outgroup members (Mikulincer & Shaver, 2001). Testing associations with both attachment dimensions would clarify the relational dynamics of contempt-proneness and help us understand the interpersonal motivations, expectations, and schemas that contempt-prone individuals bring to relationships (see also Lemay & Venaglia, in press).

Does Dispositional Contempt Predict Contempt-Related Processes in the Real World?

Finally, we examined whether dispositional contempt predicted emotional responses and relationship outcomes in the real world. We did this in three ways. First, we investigated for whom contempt-prone individuals tend to have contempt and focused on the link between dispositional contempt and attachment style in those citing contempt for someone relationally close to them. Second, we examined whether dispositional contempt predicted experiencing contempt to eliciting material, as would be expected if dispositional contempt is a product of having a lower threshold for experiencing contempt, more elicitors, higher intensity and longer-lasting episodes, and little or no ability or desire to regulate contemptuous reactions. Third, we continued probing the association of dispositional contempt and relationship functioning by examining how contemptuous people fare in their romantic relationships. The toxicity of contempt for relationships has been well-established (Gottman, 2014). We asked not only if one’s level of dispositional contempt impacted relationship outcomes but also if the perception of one’s partner as contemptuous had negative effects.

The Present Research

The goal of the present research was to look at contempt as a disposition and thus advance our understanding of contempt as an emotion. In treating dispositional contempt as a stable “affective-cognitive” (Izard, 1977) or “ideo-affective” (Tomkins, 1992) structure that drives behavior, we sought to elucidate its link to social-personality processes at various levels of analysis. First, we conducted a pilot study to develop the Dispositional Contempt Scale (DCS). This included establishing its internal and convergent validity and examining its basic demographic characteristics. In Study 1, we started to investigate the personality dynamics of dispositional contempt by assessing its associations with related tendencies, namely, other emotion dispositions (e.g., anger-proneness), the Big Five traits, and social-evaluative biases that theoretically promote contempt-proneness. Study 2 extended Study 1 by examining how dispositional contempt was related to the antisocial traits of the Dark Tetrad and more peripheral emotion dispositions like shame- and guilt-proneness. In Study 3, we focused on associations of dispositional contempt with tendencies toward social dominance vs. affiliation. Study 4 further investigated the status-differentiating and distancing functions of contempt by asking what social categories contempt targets typically fit into and whether this varies by level of dispositional contempt. Study 5 showed that dispositional contempt predicted experiencing contempt to eliciting material. Finally, in Study 6, we investigated the consequences of being a contemptuous person, particularly with regard to one’s romantic relationships.

Pilot Study

The aim of the pilot study was to develop the Dispositional Contempt Scale (DCS). Based on the features of contempt reviewed above, we created and tested an initial pool of items on two large samples, eliminated poor items, explored the factor structure of this initial measure, assessed its internal consistency, and continued to identify poor vs. strong items until we derived the 10-item DCS. Subsequently, we tested the DCS on a third sample, again tested its internal consistency, and established its convergent validity using a version of the Other as Shamer Scale (OAS; Goss, Gilbert, & Allan, 1994) that we modified to gauge the tendency to negatively view and shame others rather than feel negatively viewed and shamed by others. Finally, we explored some basic demographic characteristics of the DCS, such as whether males or females are more contempt-prone and whether contempt-proneness appears to dissipate with age, in keeping with socioemotional selectivity theory (Carstensen, 1992) and research suggesting that personality changes in adaptive ways over the life course (Roberts, Walton, & Viechtbauer, 2006).

Method

Participants and Procedure

We tested three samples, one student sample and two Mechanical Turk (MTurk) samples. MTurk is an online data collection network in which individuals participate in research studies or simple tasks for monetary compensation. MTurk has been popular in psychological research due to the large and diverse subject pool, low cost of running studies, and comparability of results to classic psychological studies (Berinsky, Huber, & Lenz, 2012; Buhrmester, Kwang, & Gosling, 2011). For Sample 1, in a mass testing session, 233 undergraduate students (66% female; median age = 21 years, range = 18–33 years) enrolled in a psychology course took our survey in exchange for course credit. The sample was ethnically representative of the university population (44% Asian, 36% Caucasian, 13% Latino, 5% African American, 1% American Indian, 1% did not report ethnic background). Sample 2 consisted of 772 MTurk workers (60% female; median age = 37 years, range = 19–74 years) who participated for monetary compensation (.25 USD for a 5-minute task). The sample was primarily Caucasian (76% Caucasian, 6% Asian, 6% Latino, 10% African American, 2% American Indian, 3% other, 4% did not report ethnic background). Sample 3 consisted of 283 MTurk workers (56% female; median age = 38 years, range = 19–71 years) who participated for monetary compensation (.50 USD for a 15-minute task) and was also primarily Caucasian (65% Caucasian, 7% Asian, 5% Latino, 9% African American, 2% American Indian, 3% other, 9% did not report ethnic background).

Participants were given the instructions, “Below are a series of statements that may or may not relate to you. Please read each statement carefully, considering each one by one, and indicate the extent to which each describes you by using the response options. There are no right or wrong answers. Please answer honestly, as we are interested in how you actually think, feel, and behave.” Participants rated the extent to which they agreed with each statement. At the end of the survey, they also provided basic demographic information.

Measures

Dispositional Contempt Scale

We created an initial item pool that was more inclusive than our conceptualization of contempt, expecting that some items would be tangential to our construct or otherwise suboptimal for tapping it (Clark & Watson, 1995). Items were written to represent cognitive, affective, and behavioral components of contempt and to include responses to morality and competence violations.¹ We avoided wording that was double-barreled or confusing and targeted reading comprehension at the 6^th grade level. We also wrote items likely to bring out individual variability in responses rather than to result in over- or under-endorsement. Twenty-five items were negatively keyed, reflecting low contemptuousness. We selected a response format with five options, ranging from 1 (Strongly disagree) to 5 (Strongly agree). All four authors wrote items, resulting in a total of 75 items.

Other as Shamer Scale – Reversed (Self as Shamer Scale)

The tendency to shame others, which is theoretically equivalent to the tendency to contemn them, was measured in Sample 3 with a modified version of the Other as Shamer Scale (OAS; Goss, Gilbert, & Allan, 1994). Its 18 items were originally designed to assess the tendency to feel shamed by others (e.g., “I feel other people see me as not good enough”). We reversed the first- vs. third-person pronouns to have this measure reflect the extent to which respondents unfavorably evaluate and look down on others rather than feel unfavorably evaluated and looked down on by others (e.g., “I see other people as not good enough”). Items were rated on a 5-point scale ranging from 1 (Never) to 5 (Almost always). Mean response on the scale was 2.41 with a standard deviation of .89 and a coefficient alpha of .94.

Results and Discussion

Developing the Dispositional Contempt Scale

To identify items that adequately represented our conceptualization of contempt, we used empirical and rational methods of good item identification. First, we utilized measures of internal consistency and eliminated items that had item-total correlations less than .35. Next we critically re-examined the content of each item and weeded out items that were long or poorly worded (e.g., “I often think the people who lack my basic competencies should know better”), inconsistent with our conceptualization of contempt (“I often feel annoyingly better than others”), and/or redundant with other items in the pool that had greater item-total correlations (e.g., “I tend not to concern myself with people who do not meet my standards”). We kept items that were face valid. After eliminating poor items, 50 items remained (see Table 1). To examine the mean level of contempt-proneness based on this item pool, we averaged responses across the 50 items, and the mean score was 2.30 with a standard deviation of .57. Scores ranged from 1.11 to 4.30, and the distribution showed a skewness of .49 (SE = .16) and kurtosis of .26 (SE = .32).

Table 1.

Item Pool for Dispositional Contempt Scale After Initial Pruning

Main Affective Process Represented	Others Tapped
Antecedent Appraisals
General Standard Violations (6 items)
Others tend to give me reasons to look down on them.
I don’t keep track of others’ shortcomings. (R)
All in all, I am repelled by others’ faults.	F, B
I tend to accept people regardless of their flaws. (R)	B
I tend to disregard people who fall short of my standards. (R)	B
I can’t stand people who don’t seem to get “the big picture.”	F
Moral Violations (13 items)
I tend to snub people who have behaved unethically.	B
I feel scorn for immoral people.	F
Selfishness is something I greatly frown upon.	B
I can still be fond of people even if they’re immoral. (R)	F
I care about everyone, even those who just look out for themselves. (R)	F, B
I would never be dismissive of someone for being a “bad person.” (R)	B
It’s easy for me to snub someone I think has acted unacceptably.	B
I can still appreciate others who don’t share my morals and values. (R)	F
I am dismissive of people who don’t have my principles.	B
People without integrity have no place in my life.	B
In general, I feel cold toward people who don’t share my values.	F
I look down on people who lack integrity.
I don’t mind people who aren’t team players. (R)	F
Competence Violations (11 items)
When I think others are incompetent, I tend to keep my distance.	B
I’m quick to roll my eyes at others’ stupidity.	B
My thoughts on others’ abilities don’t affect how well I treat them. (R)	B
I mostly avoid unintelligent individuals.	B
It doesn’t bother me when people show weakness or vulnerability. (R)	F
I generally help people who are in need of it. (R)	B
I can be rude to people who show incompetence.	B
I find myself snarling at others’ incompetent behavior.	B
I am not a harsh judge of others’ abilities. (R)
I’m rarely dismissive of others for being incompetent. (R)	B
I often feel like people don’t know what they’re doing.
Status Differentials (8 items)
I often feel like others are wasting my time.	F
I respect everyone.	F, B
I tend to discount people I see as beneath me.	B
For the most part, I respect only people who are at my level or above.	F B
I hardly ever think others are inferior to me. (R)
I often lose respect for others.	F
I’m quick to detect others’ inferiority to me.
I rarely lose esteem for others. (R)	F
Subjective Feelings/Pure Contempt (7 items)
I often have harsh thoughts about people.	A
I often feel contempt for others.
There is no one in my life I feel contempt for. (R)
I hardly ever feel disdain for others. (R)
Feeling disdain for others comes naturally to me.
I rarely feel contempt for others. (R)
At times I feel like I’m an endless well of contempt.
Action Tendencies/Behaviors (5 items)
I can be pretty insulting.
I would never try to make someone feel worthless for being who they are. (R)
I find it easy to mock people.
I treat everyone with respect, regardless of my opinion of them. (R)	A
I rarely speak badly of people even when they deserve it. (R)	A

Note. (R) indicates that the item was reverse-keyed. A = antecedent appraisal; F = subjective feeling; B = behavior/action tendency. Items in bold are in the final scale.

We administered this refined item pool to a larger sample of 772 participants to explore the psychological structure of contempt-proneness with factor analysis. If the contempt of contemptuous people tends to be elicited by detecting competence vs. moral violations, items regarding either should be differentially endorsed and form separate factors. We conducted an exploratory factor analysis on our 50 items using an unweighted least squares extraction and an oblimin rotation to examine whether contempt has a two-factor structure. We indeed extracted two factors whose eigenvalues were 16.7 and 2.6, with the first factor explaining 33.4% of the variance and the second factor, 5.2% more. However, factor loadings showed that the first factor consisted of positively keyed items, and the second factor, negatively keyed items. These factors were highly correlated, r = .64, p < .05, suggesting that these positively- and negatively-keyed items measured the same construct. Yielding two factors that consist of positively- vs. negatively-keyed items is a common finding in psychometric analyses of self-report measures (DiStefano & Morl, 2006; Marsh, 1996). Nevertheless, this two-factor structure did not reflect the expected content differentiation. Based on a scree plot indicating one large factor, results suggested a unidimensional structure. Analyzing only positively-keyed items yielded similar results. We extracted two factors that had eigenvalues of 6.8 and 1.5 and explained 32.22% and 7.34% of the variance, but these factors did not show obvious content differentiation, such as with regard to immorality vs. incompetence, and the scree plot again suggested a unidimensional structure. The coefficient alpha for all 50 items was .96, indicating high internal consistency.

Iteratively examining other factor solutions, up to 14, failed to find any that separated competence- from morality-related content, and items assessing either domain often loaded or cross-loaded onto common factors. Thus, in light of the predominantly unidimensional structure of our scale and its high coefficient alpha, we reduced its number of items to facilitate administration and eliminate redundancy. Item-total correlations on remaining items were high, so we shortened our measure by selecting items that collectively balanced having high item-total correlations and sampled content from all other domains (see Table 1) except for those dealing with morality vs. competence. This resulted in our 10-item Dispositional Contempt Scale (DCS) (see Table 2).² Three of these items were reverse-keyed. Having administered this measure to a new sample, Sample 3, we found that mean level of dispositional contempt was 2.48 with a standard deviation of .88. Internal consistency was good, with a coefficient alpha of .89. Moreover, the convergent validity of the DCS was established through its high association with the Other as Shamer Scale (Reversed), r = .85, p < .01. Finally, in terms of basic demographic characteristics, we found that level of dispositional contempt was higher in males (M = 2.73, SD = .79) than in females (M = 2.34, SD = .89, t = 3.04, p < .01), and that dispositional contempt was negatively associated with age, r = −.29, p < .001.

Table 2.

Final Dispositional Contempt Scale Items

No.	Item	Main Affective Process Represented
1	I tend to disregard people who fall short of my standards.	B, A
2	I often lose respect for others.	A/F
3	Feeling disdain for others comes naturally to me.	F
4	I tend to accept people regardless of their flaws. (R)	B, A
5	I would never try to make someone feel worthless.* (R)	B
6	I often feel like others are wasting my time.	A
7	I hardly ever think others are inferior to me. (R)	A
8	All in all, I am repelled by others’ faults.	F/B, A
9	Others tend to give me reasons to look down on them.	A
10	I often feel contempt for others.	F

Note.

^*Different version of item used in Pilot Study and Study 1, Samples 4 and 6. (R) indicates that the item was reverse-keyed. A = Antecedent appraisal; F = Subjective feeling; B = Behavior/action tendency. Response options ranged from 1 (Strongly disagree) to 5 (Strongly agree). The average of the ten items is computed to produce the scale score. In italics are words that tap the cognitive, affective, and/or behavioral components of contempt, with components labeled on the right column. Commas between components denote that they are represented by different italicized words/phrases whereas slashes indicate they are tapped by one italicized word/phrase.

In sum, the results of our pilot study allowed us to develop the DCS using rational and empirical methods and to refine our conceptualization of contempt-proneness based on its unidimensional structure. Because morality- versus competence-based items failed to create separate factors, our data suggested that morality- vs. competence-based accounts of contempt may be incomplete without each other and that contempt may be conceptualized as concerning competence or morality judgments (as well as judgments based on personal standards). A preliminary grasp of the psychology of contempt-proneness also emerged from finding that contempt is higher for individuals who are male and younger, at least once having reached adult age (no one in our sample was under 18). The gender difference was interesting, as relational aggression, which is related to contempt, is mainly used by girls and women (Underwood, 2004), and a marital interaction study found that wives expressed more contempt than husbands (Graber et al., 2011). Throughout Studies 1–6, we continued to explore the psychology of contempt-proneness while further validating the DCS, including in verifying that the DCS actually captures contemptuous experience (Study 5) and can predict important life outcomes (Study 6).

Study 1

Having developed the 10-item DCS, we proceeded to map the nomological network of dispositional contempt by examining (1) its associations with other emotion dispositions and (2) the differential associations among emotion dispositions and other personality traits. Our primary interest was in distinguishing dispositional contempt from dispositional anger and hubristic pride, which respectively tap the antagonistic and status-differentiating properties of contempt. Differentiation was expected, as anger is an approach emotion that unfolds in hopes of reparation, whereas contempt is an avoidance emotion characterized by disrespect and disregard (Fischer & Roseman, 2007; Hutcherson & Gross, 2011; Rozin et al., 1999). Moreover, despite the sense of superiority involved in both contempt and hubristic pride, that contempt is negatively-valenced and other-focused and hubristic pride, positively valenced and self-focused (Tracy & Robins, 2004), should set these dispositions apart in their relations with constructs that similarly combine these sets of features, like narcissism. We examined associations among (1) dispositional contempt, (2) related emotion dispositions, (3) the Big Five with a focus on Disagreeableness, and (4) self- and other-evaluative tendencies likely to promote contempt-proneness.

Method

Participants and Procedure

Participants came from two MTurk samples (Samples 4 and 5) and one student sample (Sample 6). In the first MTurk sample (Sample 4), 347 participants (63% female; median age = 38 years, range = 18–79 years) completed a battery of surveys in exchange for monetary compensation (1.05 USD for a 45-minute task). The sample was primarily Caucasian (83% Caucasian, 7% African American, 6% Asian American, 3% Latino, and 2% other race). In the second MTurk sample (Sample 5), 223 participants (52% female; median age = 35 years, range = 19–73 years) completed a shorter battery of surveys in exchange for monetary compensation (.50 USD for a 20-minute task). Again, this sample was primarily Caucasian (70% Caucasian, 10% African American, 8% Asian American, 6% Latino, 5% other race, and 2% did not report race). In the student sample (Sample 6), 390 undergraduate students (71% female; median age = 21 years, range = 18–61) were enrolled in a psychology course and completed a battery of surveys to fulfill a departmental prescreen requirement. This sample was ethnically representative of the university population (50% Asian, 27% Caucasian, 4% Latino, 3% African American, and 17% other race). Participants from all three samples completed the surveys online.

Due to time and space constraints, different samples were administered different measures. Of the emotion dispositions, dispositional contempt was assessed in Samples 4–6; dispositional anger, hubristic, and authentic pride, in Samples 4 and 6; dispositional disgust, in Sample 4; and dispositional envy, in Sample 5. The Big Five personality traits were measured in Samples 4–6. Finally, of the self- and other-evaluative tendencies, overt narcissism and self-esteem were assessed in Samples 4 and 6, and covert narcissism and perfectionism, in Sample 4 only. Coefficient alpha reliability and descriptive statistics for all measures are presented in Table 3.

Table 3.

Study 1: Descriptive Statistics for Emotion Dispositions, Big Five Personality Traits, and Self- and Other-Evaluative Tendencies

Measure	Sample
	4			5			6
	α	M	SD	α	M	SD	α	M	SD
DCS	.90	2.34	.92	.88	2.41	.83	.89	2.39	.70
BPAQ	.91	2.79	1.01	-	-	-	-	-	-
Anger	.87	2.56	1.23	-	-	-	-	-	-
Hostility	.88	3.04	1.43	-	-	-	-	-	-
Physical	.84	2.45	1.25	-	-	-	-	-	-
Verbal	.80	3.27	1.27	-	-	-	-	-	-
TDDS	.91	3.45	1.14	-	-	-	-	-	-
Pathogen	.85	3.77	1.21	-	-	-	-	-	-
Moral	.93	3.86	1.52	-	-	-	-	-	-
Sexual	.89	2.73	1.69	-	-	-	-	-	-
Hubristic Pride	.92	1.43	.70	-	-	-	.93	1.76	.79
Authentic Pride	.90	3.05	.94	-	-	-	.93	3.14	.75
Dispositional Envy	-	-	-	.89	1.86	.88	-	-	-
Extraversion	.88	3.00	.91	.88	2.91	.86	.83	3.14	.74
Agreeableness	.76	3.77	.72	.85	3.71	.71	.84	3.75	.63
Conscientiousness	.82	3.97	.71	.86	3.83	.69	.87	3.46	.62
Neuroticism	.83	2.70	.94	.90	2.63	.96	.89	2.94	.72
Openness	.78	3.79	.65	.84	3.55	.72	.77	3.46	.62
RSE	.91	3.92	.90	-	-	-	.92	3.66	.74
NPI	.84	1.30	.20	-	-	-	.83	1.35	.18
L/A	.78	1.36	.30	-	-	-	.78	1.42	.29
S/S	.69	1.30	.27	-	-	-	.66	1.41	.27
S/A	.49	1.28	.24	-	-	-	.45	1.27	.22
E/E	.50	1.20	.23	-	-	-	.61	1.23	.23
HNS	.84	2.82	.74	-	-	-	-	-	-
MPS	.93	3.84	.85	-	-	-	-	-	-
Self-Oriented	.93	4.50	1.23	-	-	-	-	-	-
Other-Oriented	.87	3.60	.93	-	-	-	-	-	-
Socially Prescribed	.85	3.41	.99	-	-	-	-	-	-

Note. If a statistic is not listed for a sample, the scale was not administered. RSE = Rosenberg Self-Esteem, NPI = Narcissistic Personality Inventory, L/A = Leadership/Authority, S/S = Self-Absorption/Self-Admiration, S/A = Superiority/Arrogance, E/E = Exploitativeness/Entitlement, HNS = Hypersensitive Narcissism Scale, MPS = Multidimensional Perfectionism Scale.

Measures

Dispositional Contempt

Items on the newly developed 10-item DCS (see Pilot Study) were rated on a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree).

Dispositional Anger

The Buss-Perry Aggression Questionnaire (BPAQ; Buss & Perry, 1992) is a 29-item scale that was used in Sample 4 to measure four types of aggressive tendencies: anger, hostility, physical aggression, and verbal aggression. Anger was assessed with seven items (e.g., “Some of my friends think I’m a hothead”), hostility, with eight items (e.g., “Other people always seem to get the breaks”), physical aggression, with nine items (e.g., “I have become so mad that I have broken things”), and verbal aggression, with five items (e.g., “I can’t help getting into arguments when people disagree with me”). Items were rated on a 5-point scale ranging from 1 (Extremely uncharacteristic of me) to 5 (Extremely characteristic of me). The six-item anger subscale of the Brief Affective Neuroscience Personality Scales (BANPS; Barrett, Robins, & Janata, 2013) was used to assess anger-proneness in Sample 6. Items (e.g., “When I am frustrated, I usually get angry”) were rated on a 5-point scale from 1 (Strongly disagree) to 5 (Strongly agree).

Dispositional Disgust

The Three Domain Disgust Scale (TDDS; Tybur, Lieberman, & Griskevicius, 2009) is a 21-item scale that measures individual differences in three categories of disgust sensitivity, each measured with seven items: pathogen disgust, elicited by potentially disease-causing agents (e.g., “Sitting next to someone who has red sores on their arm”); moral disgust, elicited by moral transgressions (e.g., “Forging someone’s signature on a legal document”), and sexual disgust, elicited by undesirable sexual behaviors (e.g., “Finding out that someone you don’t like has sexual fantasies about you”). Items were rated on a 7-point scale from 1 (Not at all disgusting) to 7 (Extremely disgusting).

Dispositional Authentic vs. Hubristic Pride

The Trait Pride Scale (TPS; Tracy & Robins, 2007) is a 14-item scale that measures trait-like tendencies toward authentic (achievement-based) and hubristic (global and inflated) pride. Authentic pride was assessed with seven items (e.g., “I generally feel productive”) as was hubristic pride (e.g., “I generally feel arrogant”). Items were rated on a 5-point scale from 1 (Not at all) to 5 (Extremely).

Dispositional Envy

The Dispositional Envy Scale (DES; Smith, Parrott, Diener, Hoyle, & Kim, 1999) is an 8-item scale that measures individual differences in the tendency to feel envy. A sample item is, “I feel envy every day.” Items were rated on 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree).

Big Five Personality Dimensions

The Big Five were assessed with the Big Five Inventory (BFI; John, Donahue, & Kentle, 1991), a 44-item measure that shows high convergent validity with other self-report measures of the Big Five (John et al., 2008). Extraversion was assessed with eight items (e.g., “I see myself as someone who is talkative”); Agreeableness, with nine (e.g., “I see myself as someone who is helpful and unselfish with others”); Conscientiousness, with nine (e.g., “I see myself as someone who does a thorough job”); Neuroticism, with eight (e.g., “I see myself as someone who worries a lot”); and Openness, with 10 (e.g., “I see myself as someone who is original, comes up with new ideas”). Items were rated on a 5-point scale ranging from 1 (Disagree strongly) to 5 (Agree strongly). Mean Agreeableness scores were subtracted from six to derive an index of Disagreeableness.

Perfectionism

Perfectionistic tendencies were assessed using the Multidimensional Perfectionism Scale (MPS; Hewitt & Flett, 1991), a 45-item scale that measures three facets of perfectionism: self-oriented (striving to attain perfectionism for the self), other-oriented (demanding perfectionism from close others), and socially-prescribed (striving to attain perfectionism for others). Self-oriented perfectionism was assessed with 15 items (e.g., “One of my goals is to be perfect in everything I do”), other-oriented with 15 items (e.g., “I have high expectations for the people who are important to me”), and socially prescribed with 15 items (e.g., “The people around me expect me to succeed at everything I do”). Items were rated on a 7-point scale ranging from 1 (Disagree) to 7 (Agree).

Self-Esteem

Self-esteem was assessed using the Rosenberg Self-Esteem Scale (RSE; Rosenberg, 1965), a 10-item measure designed to measure global self-esteem (e.g., “I take a positive attitude toward myself” and “I am able to do things as well as most other people”). Items were rated on a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree).

Overt Narcissism

Overt narcissism was assessed using the Narcissistic Personality Inventory (NPI; Raskin & Terry, 1988), a 40-item forced choice measure designed to measure self-aggrandizing tendencies. Participants were presented with statement pairs and asked to choose the statement that best described them. A general narcissism score and Emmons’ (1984) four subscales of leadership/authority (enjoying seeing oneself as a leader), self-absorption/self-admiration (admiring one’s personality and appearance), superiority/arrogance (thinking of oneself as grandiose), and exploitativeness/entitlement (expecting special treatment from and readily manipulating others) were computed. Narcissism was assessed with all 40 items, leadership/authority with eight (e.g., “I like having authority over people”), self-absorption/self-admiration with eight (e.g., “I like to look at myself in the mirror”), superiority/arrogance with eight (e.g., “I can usually talk my way out of anything”), and exploitativeness/entitlement with seven (e.g., “I expect a great deal from people”).

Covert Narcissism

Covert narcissism was assessed with the Hypersensitive Narcissism Scale (HSNS; Hendin & Cheek, 1997), a 10-item measure designed to measure narcissism characterized by hypersensitivity and vulnerability (e.g., “I dislike being with a group unless I know that I am appreciated by at least one of those present”). Items were rated on a 5-point scale ranging from 1 (Very uncharacteristic or untrue of me) to 5 (Very characteristic or true of me).

Results and Discussion

How Is Dispositional Contempt Associated with Other Emotion Dispositions?

Anger and Disgust

As expected given their membership in the CAD triad (Rozin et al., 1999), dispositional contempt was significantly associated with dispositional anger, its being positively associated with all four subscales of the BPAQ (see Table 4). It was highly correlated with anger, verbal aggression, and hostility, and moderately so with physical aggression. In terms of disgust, surprisingly, dispositional contempt was only weakly, and negatively, associated with its dispositional manifestation as measured with the TDDS (see Table 4). Dispositional contempt had weak and negative associations with moral and sexual disgust, and was unrelated to pathogen disgust. These results suggested that contempt-prone individuals are prone to anger, hostility, and aggression, but are somewhat less disgust-prone than those lower in dispositional contempt.

Table 4.

Study 1: How Is Dispositional Contempt Related to Other Emotion Dispositions, Including When Controlling for Disagreeableness?

Measure	r	Disagreeableness	Controlling for Disagreeableness
Buss-Perry Aggression Questionnaire	.62**	.61**	.37**
Anger	.55**	.59**	.22**
Hostility	.51**	.49**	.33**
Verbal	.55**	.59**	.22**
Physical	.45**	.40**	.35**
Three-Domain Disgust Scale	−.12*	−.15**	−.01
Pathogen	.04	.01	.08
Moral	−.16*	−.17**	−.07
Sexual	−.13*	−.15**	−.02
Hubristic Pride	.48**	.36**	.51**
Authentic Pride	−.10†	−.20**	.14
Dispositional Envy	.58**	.40**	.46**

Note. All correlations are with Sample 3, except for dispositional envy was with Sample 6.

^†p < .10;

^*p < .05;

^**p < .01.

Hubristic and Authentic Pride

As expected given their both involving feeling superior, contempt-proneness was moderately and positively associated with hubristic pride (see Table 4). Conversely, contempt-proneness was weakly and negatively associated with authentic pride, suggesting that individuals higher in dispositional contempt are prone to experiencing pride based on global, inflated self-views rather than actual achievement.

Envy

As expected in light of its status-monitoring function, dispositional contempt was strongly associated with dispositional envy (see Table 4), suggesting that people prone to experiencing contempt for those who violate important standards are also prone to experiencing envy for those successful at meeting them.

How Is Dispositional Contempt Associated with the Big Five Personality Dimensions?

As expected, across all three samples, the most robust associations were between dispositional contempt and low Agreeableness (see Table 5), or Disagreeableness, which showed no such strong association with any other emotion disposition according to Fisher r-to-z tests, all ps < .05. Dispositional contempt was also negatively associated with Extraversion, Conscientiousness, and Openness, and was positively associated with Neuroticism (see Table 5). Depending on the sample, the weakest association was with Extraversion or Openness. Because the high correlation of dispositional contempt with Disagreeableness suggested that they may reflect the same construct, we examined this statistically.

Table 5.

Study 1: How Is Dispositional Contempt, As Compared to Other Emotion Dispositions, Related to the Big Five Personality Traits?

Measure	Emotion Dispositions
	Contempt			Anger		Hubristic Pride	Envy
	4	5	6	4	6	4	5
Extraversion	−.12*	−.18**	−.09†	−.08	.02	.10†	−.29**
Agreeableness	−.78**	−.73**	−.61**	−.59**	−.52**	−.36**	.40**
Conscientiousness	−.26**	−.44**	−.30**	−.28**	−.22**	−.16**	−.51**
Neuroticism	.33**	.48**	.30**	.51**	.41**	.12*	.55**
Openness	−.11*	−.14*	−.15**	−.20*	−.13*	−.03	−.18’

Note. Numbers under emotion dispositions refer to sample.

^†p < .10;

^*p < .05;

^**p < .01

Is Dispositional Contempt Just Disagreeableness?

We fit confirmatory factor analyses (CFA) to the data, separately for Samples 4 and 6 (see Table 6 for a summary of results). First, we fit a one-factor model to the data by indicating one latent variable with the 10 DCS items and nine items from the BFI Agreeableness subscale. In each sample, the one-factor model yielded poor fit (see Table 6). We then fit a two-factor model by indicating two latent variables representing dispositional contempt and agreeableness, respectively. The two factors were allowed to covary, and the correlations between them were −.82 for Sample 4 and −.67 for Sample 6. The two-factor model yielded better fit (see Table 6). A chi-square difference test between the models was significant in both samples, and both AIC and BIC values for the two-factor solutions were less than those for the one-factor solutions. These results suggested that while the correlations between dispositional contempt and Agreeableness were large, these constructs are statistically distinct.

Table 6.

Study 1: Is Dispositional Contempt Just Disagreeableness or Dispositional Anger?

Confirmatory Factor Analyses
Dispositional Contempt vs. Disagreeableness
Sample	Factors	df	χ2	RMSEA	90% C.I.	pclosefit	CFI	TLI	AIC	BIC	Δχ2	p
4	1	152	663.708	.100	.092, .108	.000	.854	.836	16,476.53	16,621.58
	2	151	499.686	.083	.075, .091	.000	.901	.888	16,314.50	16,339.66	164.022	< .01
6	1	152	785.937	.109	.102, .117	.000	.726	.692	17,553.88	17,700.59
	2	151	592.590	.091	.084, .099	.000	.809	.784	17,362.53	17,513.10	193.347	< .01
Dispositional Contempt vs. Dispositional Anger
Sample	Factors	df	χ2	RMSEA	90% C.I.	pclosefit	CFI	TLI	AIC	BIC	Δχ2	p
4	1	119	1,109.194	.156	.148, .165	.000	.722	.683	17,314.45	17,444.74
	2	118	398.840	.084	.075, .093	.000	.921	.909	16,606.10	16,740.22	710.350	< .01
6	1	90	580.869	.123	.114, .133	.000	.738	.694	14,816.09	14,932.67
	2	89	356.112	.091	.081, .101	.000	.857	.831	14,593.33	14,713.80	224.757	< .01

Note. RMSEA = Root mean square error of approximation. C.I. = Confidence interval. CFI = Comparative fit index. TLI = Tucker-Lewis Index. AIC = Aikake information criterion. BIC = Bayesian information criterion.

To further examine whether dispositional contempt and Agreeableness were distinct and also assess their content differentiation, we attempted to replicate our confirmatory factor model using a data-driven EFA for Samples 4 and 6 (see Table 7). We specified the extraction of two factors and used oblimin rotation to enhance interpretation of loadings. Measures of EFA model fit for Sample 4 (TLI = .903 & RMSEA = .077, 90% CI [.067, .085]) and 6 (TLI = .822, RMSEA = .083, 90% CI [.073, .09]) were similar to those observed for the two-factor CFA model. Overall, dispositional contempt and Agreeableness showed clear loading patterns consistent with theoretical expectations and the specifications from our CFA models. In further analyses (see also Table 4), we controlled for Disagreeableness to continue to explore the unique explanatory power of dispositional contempt. We also replicated the CFAs with contempt- vs. anger-proneness to ensure that they, too, were statistically distinguishable.

Table 7.

Study 1: Exploratory Factor Analysis Loading Patterns for Dispositional Contempt vs. Agreeableness Items on Two Factors

Item	Sample
	4		6
	F1	F2	F1	F2
Is considerate and kind to almost everyone	.02	.82	.00	.81
Is helpful and unselfish with others	.05	.69	.01	.65
Has a forgiving nature	−.01	.68	−.05	.55
Likes to cooperate with others	−.12	.56	−.03	.54
Is generally trusting	−.09	.54	.08	.60
Is sometimes rude to others (R)	−.26	.47	−.26	.30
Can be cold and aloof (R)	−.27	.35	−.34	.19
Starts quarrels with others (R)	−.43	.18	−.31	.32
Tends to find fault with others (R)	−.52	.16	−.49	−.01
Feeling disdain for others comes naturally to me.	.85	.00	.71	−.04
I often feel contempt for others.	.86	.16	.54	.05
All in all, I am repelled by others’ faults.	.84	−.01	.75	.03
Others tend to give me reasons to look down on them.	.82	.01	.77	.08
I tend to disregard people who fall short of my standards.	.75	−.06	.72	−.13
I often feel like others are wasting my time.	.69	−.04	.66	−.02
I often lose respect for others.	.62	−.11	.63	.00
I would never try to make someone feel worthless. (R)	.73	−.01	.68	.05
I hardly ever think others are inferior to me. (R)	.50	−.12	.09	−.23
I tend to accept people regardless of their flaws. (R)	.45	−.33	.20	−.38

Note. (R) denotes items were reverse-keyed. F1 and F2 represent dispositional contempt and Agreeableness, respectively. Bolded loadings indicate correspondence between an indicator and its factor. EFA for Sample 4 and 6 used oblimin rotation. Fit for the EFA two-factor model for Samples 4 and 6 were comparable to those observed in our corresponding CFA two-factor models. Overall, items demonstrated good discrimination between their factors and were fairly consistent between samples.

Is Dispositional Contempt Just Dispositional Anger?

Similar to our comparisons of dispositional contempt and Disagreeableness, we fit separate CFAs to Samples 4 and 6. First, we fit a one-factor model by indicating one latent variable with the 10 DCS items and, for Samples 4 and 6, respectively, the seven items from the BPAQ anger subscale and six items from the BANPS anger subscale. The one-factor model yielded poor fit in both samples (see Table 6). We then fit a two-factor model by indicating two latent variables representing dispositional contempt and dispositional anger, respectively. The two factors were allowed to covary, and the correlation between them was .57 and .50 in Samples 4 and 6. The two-factor model yielded better fit (see Table 6). A chi-square difference test between the models was significant in each sample, and both AIC and BIC values for the two-factor solutions were less than corresponding values for the one-factor solutions. As in analyses with Disagreeableness, these results suggested that while correlations of dispositional contempt and anger were large, they, too, are relatively distinct.

What Self- and Other Evaluative Tendencies Are Associated With Dispositional Contempt?

As expected, dispositional contempt, which theoretically involves the consistent appraisal of others as falling short of one’s standards, was significantly and positively associated with perfectionism, including with each of its variants as tapped by the MPS. It was most highly associated with other-oriented perfectionism, followed by socially prescribed perfectionism and self-oriented perfectionism. This set of results held when controlling for Disagreeableness (see Table 8). According to Fisher r-to-z tests, compared to dispositional anger and hubristic pride, dispositional contempt was more strongly associated with perfectionism and its variants in all cases except for that hubristic pride was similarly related to self-oriented perfectionism. Their association disappeared when controlling for dispositional contempt. This set of results is in keeping with conceptualizing contempt as primarily other-directed (expectations for others, others’ expectations for oneself) with reference to important standards.

Table 8.

Study 1: How Are Self-/Other-Evaluative Tendencies Associated With Dispositional Contempt, Compared to Other Emotion Dispositions and When Controlling for Disagreeableness?

Measure	Emotion Dispositions			Disagreeableness	Controlling for Disagreeableness
	Contempt	Anger	Hubristic Pride
MPS	.43**	.22**	.31**	.31**	.47**
Self-Oriented	.20**	.06	.16**	.10*	.31**
Other-Oriented	.48**	.25**	.31**	.39**	.46**
Socially Prescribed	.41**	.26**	.31**	.32**	.40**
RSE	−.25**	−.29**	−.04	−.31**	−.02
NPI	.28**	.19**	.40**	.15**	.43**
L/A	.16**	.07	.28**	.05	.30**
S/S	.16**	.12*	.34**	.08†	.26**
S/A	.22**	.12*	.33**	.12*	.33**
E/E	.51**	.35**	.44**	.38**	.54**
HNS	.59**	.45**	.32**	.52**	.48**

Note. These correlations are for Sample 4. MPS = Multidimensional Perfectionism Scale. RSE = Rosenberg Self-Esteem, NPI = Narcissistic Personality Inventory, L/A = Leadership/Authority, S/S = Self-Absorption/Self-Admiration, S/A = Superiority/Arrogance, E/E = Exploitativeness/Entitlement, HNS = Hypersensitive Narcissism Scale.

^†p < .10;

^*p < .05;

^**p < .01.

Finally, perhaps suggesting a role for defensive self-esteem regulation, dispositional contempt, despite involving looking down on others, was associated with low self-esteem (see Table 8). Yet, dispositional contempt was associated with narcissism. These respective associations increased in magnitude when self-esteem and narcissism were unconfounded from each other (see Paulhus et al., 2010). In Sample 4, for example, dispositional contempt was associated with self-aggrandizing narcissistic tendencies, r = .38, p < .001, and with genuine feelings of self-worth, r = −.35, p < .001.³ In terms of the four facets of narcissism, dispositional contempt was associated with all four of them in Sample 4 and with three of four in Sample 6 (see Table 8). It was consistently most highly associated with the most toxic and maladaptive facet of narcissism, exploitativeness/entitlement (Emmons, 1987; Raskin & Novacek, 1989), more so than were the other emotion dispositions. Finally, consistent with its link to low self-esteem, dispositional contempt was strongly associated with hypersensitive narcissism. This association was significantly higher than that between dispositional contempt and grandiose narcissism, z = 5.11, p < .001, highlighting the fragile, if inflated, self-views of contemptuous people. Except for the association with low self-esteem, the above associations remained significant when controlling for Disagreeableness.

Thus, results of Study 1 largely supported our hypotheses and suggested that the contemptuous person tends to be dispositionally angry, envious, and hubristic, but not dispositionally disgusted, not even morally disgusted. Part of what distinguished contempt-proneness from other emotion dispositions was its robust association with Disagreeableness. Still, dispositional contempt was more than just Disagreeableness, as demonstrated in factor analyses and the personality correlates of these constructs. Overall, we found that contempt-proneness is characterized by a dual nature. On the one hand, on top of being disagreeable, the contempt-prone person was seen to be narcissistic and to have perfectionistic standards for others. On the other hand, he or she was shown to be susceptible to the vulnerable form of narcissism and to feel that others impose perfectionistic standards on him or her. Thus, the tendency to see and treat others as defective might paradoxically involve having inflated but negative self-views. In light of the association between dispositional contempt and socially toxic variables like Disagreeableness and entitlement/exploitativeness, we wished to further characterize the antisocial nature of contempt-proneness and did so in Study 2.

Study 2

The purpose of Study 2 was two-fold. First, having observed the particularly strong association of dispositional contempt with Disagreeableness and the exploitativeness/entitlement facet of narcissism, we sought to determine how dispositional contempt was related to other personality traits underpinned by antisociality, specifically, the Dark Tetrad of narcissism (also examined in Study 1), Machiavellianism (the tendency to manipulate and exploit others), psychopathy (callous disregard for others alongside tendencies toward impulsive sensation-seeking), and sadism (taking satisfaction in others’ pain and misfortunes) (Buckels, Jones, & Paulhus, 2013; Chabrol et al., 2009; Furnham et al., 2013). We expected dispositional contempt to be associated with each Dark Tetrad trait and also examined how they (1) uniquely contributed to dispositional contempt and (2) were associated with hubristic pride, which, relative to dispositional contempt, had a similar pattern of Dark Tetrad correlates but was more highly associated with narcissism in Study 1.

Second, we wanted to supplement the nomological network of emotion dispositions initiated in Study 1 with the more distally related emotions of shame and guilt, with predictions that dispositional contempt would be related to shame-proneness. Unlike guilt and anger, which are behavior-focused, contempt and shame are rooted in the appraisal of the global person (Fischer & Roseman, 2007; Lewis, 2008). Both shame and contempt involve deeming an entire person – whether self or other – grossly unacceptable, worthless, or inferior in light of a wrongdoing, and both involve the action tendency to keep that person – whether self or other – away, such as through hiding in the case of shame, or rejection and avoidance in the case of contempt. These attributional similarities do not necessitate an association between contempt- and shame-proneness but a basis for one comes from having observed in Study 1 the association of dispositional contempt with low self-esteem, hypersensitive narcissism, and socially-prescribed perfectionism, all of which are maladaptive and predispose individuals to perceiving their shortcomings.

Method

Participants and Procedure

Participants were 290 MTurk workers (Sample 7) who completed a battery of surveys in exchange for monetary compensation (1.05 USD for a forty five minute task). This sample was primarily female (61% female, 39% male), had a median age of 33 years (range = 18– 83 years), and was largely Caucasian (82% Caucasian, 7% African American, 5% Asian American, 3% Latino, 1% American Indian, and 2% other ethnicity). Participants completed surveys online. Coefficient alpha reliability and descriptive statistics for all measures are presented in Table 9.

Table 9.

Study 2: Descriptive Statistics for Emotion Dispositions and Dark Tetrad Personality Traits

Measure	α	M	SD
Emotion Dispositions
Contempt	.91	2.39	.85
Hubristic Pride	.91	1.61	.82
Shame	.92	2.00	.95
Guilt	.90	2.29	1.00
Dark Tetrad
Narcissism	.79	2.68	.69
Machiavellianism	.81	2.80	.63
Psychopathy	.82	2.01	.72
Sadism	.72	1.61	1.61

Measures

Dispositional Contempt

As in Study 1, the Dispositional Contempt Scale was used to assess level of dispositional contempt.

Dispositional Hubristic Pride

As in Study 1, the Trait Pride Scale was used to assess level of dispositional hubristic pride.

Dark Tetrad

The Dark Tetrad personality traits were assessed with two measures. Narcissism, Machiavellianism, and psychopathy were measured with the 28-item Short Dark Triad (SDT; Jones & Paulhus, 2014), on which narcissism was assessed with nine items (e.g,. “I insist on getting the respect I deserve”), Machiavellianism, with 10 items (e.g., “It’s wise to keep track of information that you can use against people later”), and psychopathy, with nine items (e.g., “It’s true that I can be cruel”). Items were rated on a 5-point scale ranging from 1 (Disagree strongly) to 5 (Agree strongly). Sadism was assessed with the Short Sadistic Impulse Scale (SSIS; O’Meara, Davies, & Hammond, 2011), a 10-item scale designed to measure the tendency to enjoy asserting dominance and inflicting harm on others (e.g., “I enjoy seeing people hurt”). Items were rated on a 5-point scale ranging from 1 (Unlike me) to 5 (Like me).

Dispositional Guilt and Shame

Dispositional tendencies toward shame and guilt were assessed using a modified version of the State Shame and Guilt Scale (SSGS; Marschall, Saftner, & Tangney, 1994), a 15-item scale originally designed to assess momentary experiences of shame (feelings of worthlessness about the global self), guilt (feelings of remorse about something bad one has done), and pride (feelings of achievement and competence). To assess proneness to these emotions, instructions were modified from “Rate each statement based on how you are feeling right at this moment” to “Rate each statement based on how you generally feel (i.e., on average).” Dispositional shame was assessed with five items (e.g., “I generally feel worthless, powerless”), as was dispositional guilt (e.g., “I generally feel remorse, regret”). Items were rated on a 5-point scale ranging from 1 (Not at all) to 5 (Extremely).

Results

How Is Dispositional Contempt Associated with the Dark Tetrad Personality Traits?

Consistent with Study 1, dispositional contempt was significantly positively associated with narcissism. To a greater extent (see Table 10), contempt was positively associated with Machiavellianism, psychopathy, and sadism, indicating that contempt-prone individuals are antisocial along each Dark Tetrad dimension. To determine which dimensions of the Dark Tetrad were uniquely related to dispositional contempt, we ran a multiple regression model simultaneously entering all four Dark Tetrad traits in the prediction of dispositional contempt. The model (R² = .56, F = 89.44, p < .001) showed that dispositional contempt was primarily explained by Machiavellianism (B = .43, p < .001), followed by psychopathy (B = .36, p < .001) and sadism (B = .11, p < .05), but not narcissism. Further implicating a strong link with Machiavellianism, comparing the magnitude of the Dark Tetrad correlates of dispositional contempt vs. hubris showed that Machiavellianism was more strongly related to dispositional contempt than to hubris, z = 3.39, p < .001. This finding is interesting because hubristic pride is intrinsically linked to status differentials through a sense of superiority to others, but it may be that contempt-proneness involves more status-sensitivity, status-monitoring, and/or readiness to manipulate others to achieve status.

Table 10.

Study 2: How Is Dispositional Contempt Related to the “Dark Tetrad” Personality Traits and Dispositions Toward Shame and Guilt?

	Emotion Dispositions				Dark Tetrad Personality Traits
	1	2	3	4	5	6	7	8
1. Contempt	-
2. Hubristic Pride	.57**	-
3. Shame	.42**	.38**	-
4. Guilt	.35**	.33**	.76**	-
5. Narcissism	.26**	.34**	−.14*	−.12*	-
6. Machiavellianism	.66**	.47**	.34**	.32**	.25**	-
7. Psychopathy	.64**	.59**	.34**	.31**	.35**	.55**	-
8. Sadism	.41**	.48**	.32**	.26**	.16**	.32**	.45**	-

Note.

^*p < .05;

^**p < .01

To more directly examine the association of Machiavellianism with dispositional contempt vs. hubris, we ran a regression model that predicted Machiavellianism from both of these emotion dispositions. The model (R² = .44, F = 113.570, p < .001) showed that both dispositional contempt (B = .59, p < .001) and hubristic pride (B = .12, p < .05) were unique predictors of Machiavellianism, but dispositional contempt was to a greater extent. Dispositional contempt and hubris were unique predictors of the remaining Dark Tetrad traits except in the case of narcissism (R² = .12, F = 19.869, p < .001), which was no longer associated with dispositional contempt (B = .08, ns) when controlling for hubristic pride (B = .29, p < .001). Consistent with a greater role for narcissism in hubristic pride, a multiple regression analysis that predicted hubristic pride from the Dark Tetrad (R² = .45, F = 57.93, p < .001) showed unique associations with each including narcissism (B = .14; then, in decreasing order: psychopathy, B = .34; sadism, B = .25; and Machiavellianism, B = .17; all p’s < .001).

How is Dispositional Contempt Associated with Dispositional Shame and Guilt?

Dispositional contempt was moderately positively associated with dispositional shame, indicating that individuals with higher levels of dispositional contempt were more shame-prone than those with lower levels (see Table 10). Although dispositional contempt was also associated with guilt-proneness, when shame- and guilt-proneness were unconfounded from each other (Paulhus et al., 2010), only the association with shame-proneness remained, r = .25, p < .001. Results held when controlling for hubristic pride, which itself was uniquely if less strongly associated with both emotion dispositions. Thus, what emerged is a portrait of the contemptuous individual as antisocial as well as shame-prone. Consistent with these results, guilt-proneness, which has been established in past research as associated with moral behavior (Cohen, Panter, & Turan, 2012), was not uniquely related to dispositional contempt.

Study 3

That dispositional contempt operates within the context of what has been called the dominance behavioral system (Johnson, Leedom, & Muhtadie, 2012), rank regulation system (Zuroff, Fournier, Patall, & Leybman, 2010), hierarchical domain (Bugental, 2000), or power system (Shaver, Segev, & Mikulincer, 2011) was established in Study 2 through its strong link to Machiavellianism. Given the apparent concern of the contemptuous person with getting ahead as opposed to getting along, our goal in Study 3 was to continue to examine the association of dispositional contempt with the interpersonal axes of social power versus affiliation (Wiggins, 1979). Building on the link between dispositional contempt and Machiavellianism, we examined the power motives of contempt-prone individuals in intergroup contexts. Contemptuous individuals likely believe that entire groups, not just individuals, are inferior to others, a perspective captured by social dominance orientation. In a related vein, past work established hubristic pride as linked to racism (Ashton-James & Tracy, 2012), and we predicted that dispositional contempt would be even more strongly associated with racism given its direct tie to derogating others.

Whereas the power motive involves the desire to influence, dominate, or obtain control over others or resources, the affiliation motive reflects a drive to establish, maintain, and restore positive relations with others. We examined how dispositional contempt was associated with three tendencies that reflect the desire to be close to others – attachment style, need to belong, and loneliness. We expected contempt-prone individuals to be characterized by attachment avoidance as suggested by their social aversion and the use of contempt as a social distancing mechanism (Fischer & Roseman, 2007). However, we also considered the possibility that contempt-proneness would be associated with attachment anxiety to the extent that contempt is a reaction to others’ falling short of one’s expectations, which may extend to one’s relational needs and their feeling unmet, possibly because they are too high. Furthermore, as suggested by links with hypersensitive narcissism and socially-prescribed perfectionism in Study 1, contempt-prone individuals may themselves worry about falling short, being rejected and abandoned, including by their partners, all suggestive of attachment anxiety (Hazan & Shaver, 1994). In either case, we expected the contempt-prone to report weak belonging needs but to be lonely.

Method

Participants and Procedure

Participants came from two student samples (Samples 8 and 9). In Sample 8, 1,356 undergraduate students enrolled in a psychology course completed a battery of surveys to fulfill a departmental prescreen requirement. This sample was primarily female (68%) with a median age of 21 years (range = 18– 61 years) and was ethnically representative of the university population (50% Asian, 27% Caucasian, 4% Latino, 3% African American, 1% American Indian, and 15% other ethnicity). Similarly, in Sample 9, 1,109 participants completed a battery of surveys to fulfill a departmental prescreen requirement. This sample was primarily female (66% female) with a median age of 19 years (range = 18–59 years) and was ethnically representative of the university population (41% Asian American, 34% Caucasian, 21% Latino, 3% African American, <1% other ethnicity, <1% did not report ethnicity). Coefficient alpha reliabilities and descriptive statistics for all measures are presented in Table 10.

Measures

Dispositional Contempt

As in Studies 1 and 2, the Dispositional Contempt Scale was used to assess level of dispositional contempt.

Dispositional Hubristic Pride

As in Studies 1 and 2, the Trait Pride Scale was used to assess level of dispositional hubristic pride.

Social Dominance Orientation

The tendency to believe that some groups are inherently better than others was measured with the Social Dominance Orientation Scale (SDO; Pratto, Sidanius, Stallworth, & Malle, 1994), a measure of 14 items (e.g., “Some people are just more worthy than others.”) rated on a 7-point scale from 1 (Strongly agree) to 7 (Strongly disagree).

Motivation to Control Prejudice

Intrinsic versus extrinsic motivation to control prejudice toward African-Americans was measured using the Internal and External Motivation to Control Prejudice Scale (Plant & Devine, 1998), a 10-item scale with two subscales: internal motivation (IMS; e.g., “Because of my personal values, I believe that using stereotypes about Black people is wrong.”) and external motivation (EMS; e.g., “”I try to act nonprejudiced toward Black people because of pressure from others.”). Items were rated on a 9-point scale from 1 (Strongly agree) to 9 (Strongly disagree).

Racism

Racism toward African Americans was measured using the Attitudes Toward Blacks Scale (ATB; Brigham, 1993), a 20-item scale with items (e.g., “Black people are demanding too much too fast in their push for equal rights.”) rated on a 7-point scale from 1 (Strongly disagree) to 7 (Strongly agree).

Adult Attachment Style

Adult attachment style was assessed using the Experiences in Close Relationships-Revised Questionnaire (ECR-R; Fraley, Waller, & Brennan, 2000), a 36-item measure designed to assess adult attachment orientations of avoidance (being uncomfortable with being close to romantic partners) and anxiety (insecurity about the availability and responsiveness of romantic partners). Avoidance was assessed with 18 items (e.g., “I get uncomfortable when a romantic partner wants to be very close”), and anxiety, with 18 items (e.g., “I often worry that my partner doesn’t really love me”). Items were rated on a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree).

Belonging Needs

Belonging needs were measured with the Need to Belong Scale (NBS; (Leary, Kelly, Cottrell, & Schreindorfer, 2006), a 10-item measure designed to gauge individuals’ need to belong (e.g., “I try hard not to do things that will make other people avoid or reject me” and “It bothers me a great deal when I am not included in other people’s plans”). Items were rated on a 5-point scale ranging from 1 (Strongly disagree) to 5 (Strongly agree).

Loneliness

Loneliness was assessed using the UCLA Loneliness Scale (Version 3) (UCLA-3; Russell, 1996), a 20-item measure designed to measure loneliness in social relationships (e.g., “How often do you feel that your interests and ideas are not shared by those around you?” and “How often do you feel that no one really knows you well?”). Items were rated on a 4-point scale ranging from 1 (Never) to 4 (Always).

Results

Is Dispositional Contempt Associated With Social Dominance?

As shown in Table 12, dispositional contempt was significantly associated with having a higher social dominance orientation, greater motivation to control prejudice for external factors but lower motivation to control it for internal factors, and harboring more negative attitudes toward African Americans, i.e., racism. In all cases except for with regard to external motivation to control prejudice, the magnitude of these correlations was greater than those observed with hubristic pride according to Fisher-r-to-z tests, all ps < .001. Thus, the other-derogation inherent to contempt robustly manifested at the level of outgroup derogation. A series of regression analyses showed that dispositional contempt and hubristic pride were unique predictors of all of these dominance variables except for internal motivation to control prejudice, which was no longer associated with hubristic pride after controlling for dispositional contempt (see Table 12 for all standardized betas).

Table 12.

Study 3: How Is Dispositional Contempt as Compared to Hubristic Pride Related to Social Dominance and Affiliation?

Measure	Dispositional Contempt				Dispositional Hubristic Pride
	8		9		8		9
	r	B	r	B	r	B	r	B
Social Dominance
Social Dominance Orientation	.29**	.24**	-	-	.12**	.12**	-
External Motivation to Control Prejudice	.16**	.11**	-	-	.15**	.11**	-
Internal Motivation to Control Prejudice	−.22**	−.22**	-	-	−.14**	−.04	-
Attitudes Toward Blacks	-	-	.43**	.38**	-		.28**	.13**
Social Affiliation
Avoidant Attachment	.25**	.27**	.19**	.19**	.06*	−.05†	.08*	.01
Anxious Attachment	.30**	.29**	.24**	.22**	.14*	.03	.14**	.05
Belonging Needs	.10**	.11**	−.00	.00	.02	−.03	−.01	−.01
Loneliness	.35**	.39**	-	-	.18**	−.09	-	-

Note. Numbers under emotion dispositions refer to sample. If a statistic is not listed for a sample, the scale was not administered. B = standardized regression coefficients derived from multiple regression models in which dispositional contempt and hubristic pride were entered simultaneously in the prediction of these social dominance vs. affiliation variables. Thus, associations with one emotion disposition controlling for the other are shown.

^†p < .10;

^*p < .05;

^**p < .01.

Is Dispositional Contempt Associated With Social Disconnection?

As shown in Table 12, associations with the ECR-R indicated that contemptuous individuals tended to be insecurely attached. As predicted, dispositional contempt was positively associated with attachment avoidance and anxiety. According to Fisher r-to-z tests, both ps < .05, both associations were greater than the associations found between attachment and hubristic pride. Furthermore, only dispositional contempt was a unique predictor of either attachment dimension when dispositional contempt and hubris were used to control for each other in regression models (see Table 12 for standardized betas). This suggests that only dispositional contempt was uniquely associated with attachment style. Conversely, when avoidance and anxiety were used to predict dispositional contempt, both were unique predictors, with roughly equal contributions from anxiety (Bs = .27 and .21 in Samples 8 and 9) and avoidance (Bs = .20 and .14). Finally, depending on the sample, dispositional contempt was either weakly positively related or unrelated to need to belong and was associated with loneliness, significantly more so than was hubristic pride (see Table 11).

Table 11.

Study 3: Descriptive Statistics for Emotion Dispositions Social Dominance and Affiliation Variables

Measure	Sample
	8			9
	α	M	SD	α	M	SD
Emotion Disposition
Contempt	.87	2.17	.62	.90	2.27	.62
Hubristic Pride	.88	1.57	.60	.90	1.61	.64
Social Dominance
Social Dominance Orientation	.95	2.29	1.15	-	-	-
External Motivation to Control Prejudice	.81	4.70	1.86	-	-	-
Internal Motivation to Control Prejudice	.84	6.18	1.20	-	-	-
Attitudes Toward Blacks	-	-	-	.86	2.48	.80
Interpersonal Connectedness
Avoidant Attachment	.90	3.30	.96	.91	3.33	1.00
Anxious Attachment	.92	3.87	1.07	.91	3.93	1.07
Belonging Needs	.81	3.39	.65	.80	3.37	.63
Loneliness	.92	1.95	.53	-	-	-

Note: If a statistic is not listed for a sample, the scale was not administered.

Taken together, results from Study 3 suggested that contempt-prone individuals are motivated by the need to dominate and derogate others, including at the intergroup level. Nevertheless, in keeping with the paradoxical nature of dispositional contempt, contempt-prone individuals showed a slight, if any, need to belong and a tendency toward attachment anxiety as well as attachment avoidance. He or she was also lonely, an outcome that could be facilitated by the accumulating impact he or she has on others. Using hubristic pride as a point of comparison, we demonstrated that problems along the affiliation dimension are more implicated in dispositional contempt, and that, ultimately, interpersonal antagonism amidst interpersonal fragility is a more relevant dynamic for understanding dispositional contempt than dispositional hubris. Because adult attachment primarily speaks to interpersonal dynamics within the context of a romantic relationship, one of our aims in Study 4 was to examine for whom contemptuous people tend to have contempt and how attachment is related to contempt-proneness in individuals citing contempt for their romantic partners.

Study 4

To further explore the status-differentiating and social distancing functions of contempt, whereby others get stepped on in one’s attempts to get ahead or kept at arm’s length in the face of dwindling expectations, we investigated for whom individuals – especially contempt-prone individuals – tend to have contempt. We asked participants to nominate someone in their lives for whom they have contempt and examined the social categories to which these targets belonged. To gain specificity with regard to contempt, we compared typical contempt targets to anger targets. If contempt is a status differentiator, regulator of group resources, and distancing mechanism, we predicted that contempt targets would tend to be at participants’ same level of the status hierarchy, come from settings in which group processes are likely hindered (e.g., at work), and/or be people with whom participants do not feel pressure to maintain positive relationships. Conversely, because anger helps repair relationships over the long run (Fischer & Roseman, 2007), we predicted that typical anger targets would be relationally closer to participants and be more likely to be invested in. Along these lines of reasoning, we thought it would take a greater level of dispositional contempt to nominate a contempt target to which one was close. To better understand the underlying dynamics of such a scenario, we assessed whether attachment avoidance and anxiety were differentially linked to dispositional contempt in participants who voluntarily nominated romantic partners, past or present, as contempt targets.

Method

Participants and Procedure

Participants were 477 MTurk workers (Sample 10) who completed a battery of surveys in exchange for monetary compensation (.50 USD for a 10-minute task). This sample was balanced in terms of gender (52% female) with a median age of 35 years (range = 18– 84 years) and was largely Caucasian (85% Caucasian, 5% African American, 5% Asian American, 3% Latino, 1% American Indian, and 2% other ethnicity). Participants completed surveys online.

Adult Attachment Style

As in Study 3, attachment style was assessed using the ECR-R (Fraley, Waller, & Brennan, 2000) with avoidance having a M = 3.16, SD = .86, and anxiety, M = 2.58, SD = .90.

Nominated Contempt and Anger Targets

Participants each provided one contempt target and one anger target, in counterbalanced order, and were asked to explain why these individuals prompted these emotions and to indicate the relationship had with these individuals. Parallel prompts were provided for soliciting target nominations for contempt vs. anger based on work by Hutcherson and Gross (2011) and Fischer and Roseman (2007) that distinguished the cognitive appraisals and action tendencies related to these emotions. Appraisals giving rise to anger, for example, involved wanting wrongs to be corrected, whereas those preceding contempt, apprehension that standards would be met. The prompts were as follows:

• “Please think about someone you know in ‘real life’ that you have contempt for – someone you have a low opinion of, feel ‘cold’ toward, can’t stand, and generally want nothing to do with. Briefly explain why you’ve felt that way about them.”

• “Please think about someone you know in ‘real life’ that’s made you angry – someone who did something that had a negative impact on you, that you felt ‘hot’ toward, and that’s needed to make reparations. Briefly explain why you’ve felt that way about them.”

Notably, these prompts presented somewhat of a trait/state confound in that contempt was presented as being about the target in general and to have lingered longer, whereas anger involved more of a situational occurrence focused on the target’s behavior, but these differences are in keeping with distinctions made between these two emotions in the aforementioned empirical work (Fischer & Roseman, 2007; Hutcherson & Gross, 2011). For example, Fischer & Roseman (2007) found that contempt, relative to anger, encourages focus on others’ dispositions rather than their isolated behaviors and leads to more stable and enduring attributions, including permanent changes in beliefs about others.

After making and describing the circumstances behind their nominations, participants were asked, “How would you describe this person’s relationship to you?” and were provided with the response options “acquaintance,” “boyfriend/girlfriend,” “child,” “co-worker,” “employee,” “ex-boyfriend/girlfriend,” “ex-spouse,” “extended family (e.g., cousin, aunt),” “friend,” “in-law,” “parent,” “sibling,” “spouse” and “other:” followed by a free-response answer box. Short answer explanations and free-response labels were each read to ensure that all categories were covered in our coding scheme. “Boss” and “neighbor” are examples of two that were added. Subsequently, responses were coded into the categories seen in Figure 1.

Figure 1.

Percentage breakdown (n = 477) of type of relationship that contempt targets (left bar) and anger targets (right bar) had with participants who nominated them in Study 4. Each participant nominated both a contempt and anger target. For ease of interpretation, relation type is shown in decreasing order of representation by contempt targets.

Results and Discussion

For Whom Do Individuals Tend to Have Contempt?

As shown in Figure 1, the majority of participants nominated contempt targets who could be categorized in one of two ways: (1) as at the same level of the status hierarchy as participants or, probably more accurately and instructively, (2) as relationally distant to the participant. Even within the overarching category of family, contempt targets tended to be part of participants’ extended, not immediate, family. Acquaintances, co-workers, and extended family were the most common types of relationship observed and together comprised 54.8% of contempt targets. This finding is interesting because contempt is a distancing emotion and yet these targets were characterized by distance. For example, one might have expected more relationally close targets to be nominated based on their greater ability and opportunity to elicit strong emotional experiences like contempt.

Alternatively, because acquaintances and co-workers may be more accessible in mind as sources of any kind of emotion given their greater number and/or prevalence in everyday life, it was important to examine the specificity of the above findings to contempt by comparing the social categories of typical contempt targets to those of anger targets (see Figure 1). Here, a reversal was observed. The majority of participants nominated anger targets who were relationally close to the participant. Romantic partners, immediate family, and friends were the most common relationship type observed and together comprised 51.0% of identified anger targets. In fact, about twice as many anger targets, relative to contempt targets, were romantic partners, immediate family, or friends, and about twice as many contempt targets, relative to anger targets, were acquaintances or extended family.

For Whom Do The Contemptuous Tend to Have Contempt?

To examine whether dispositional contempt was associated with the social categories to which nominated contempt targets belonged, we tested for differences in mean levels of dispositional contempt observed in each group (see Table 13). Group differences only emerged for the smallest groups at the extreme ends. To have more power to detect differences, we combined acquaintances, co-workers, and extended family into one category of “distal targets,” and romantic partners, immediate family, and friends into another category of “close targets,” then tested for differences in mean levels of dispositional contempt between individuals identifying targets in either group. Counter to expectation, individuals selecting close contempt targets (M = 2.30, SD = .81) did not have higher levels of dispositional contempt than those selecting distal contempt targets (M = 2.23, SD = .79, t(376) = .681, ns), although differences were in expected directions. A similar analysis performed for anger targets showed no differences in dispositional contempt between individuals selecting distal (M = 2.16, SD = .82) vs. close (M = 2.24, SD = .76, t(376) = 1.11, ns) anger targets.

Table 13.

Study 4: How Does Level of Dispositional Contempt Differ by Social Categories of Nominated Contempt vs. Anger Targets?

Contempt Target	n	Dispositional Contempt		Anger Target	n	Dispositional Contempt
		M	SD			M	SD
Group	5	2.52	.75	Stranger	4	2.68	.53
Friend	45	2.35	.76	Employee	12	2.40	1.24
Immediate Family	38	2.32	.89	No One	6	2.35	1.08
Employee	10	2.29	1.26	Immediate Family	79	2.34	.79
Extended Family	63	2.26	.79	Boss	21	2.33	.81
Neighbor	19	2.25	.87	Group	3	2.33	.91
Co-Worker	88	2.24	.84	Romantic Partner	81	2.26	.67
Romantic Partner	45	2.23	.80	Acquaintance	53	2.23	.86
Acquaintance	99	2.22	.75	Service	6	2.20	1.05
Boss	22	2.18	.73	Roommate	5	2.20	1.48
No One	11	2.06	.99	Neighbor	9	2.18	.78
Roommate	4	1.63	.67	Extended Family	35	2.14	.80
Public Figure	3	1.63	.25	Friend	76	2.13	.81
Service	5	1.59	.60	Co-Worker	72	2.10	.81
Stranger	-	-	-	Public Figure	1	1.8	-

Note. Contempt and anger target categories are respectively presented in order of decreasing level of mean dispositional contempt for each category.

How Does Attachment Contribute to Dispositional Contempt in Individuals Who Nominated Romantic Partners as Contempt Targets?

Selecting a romantic partner as a contempt target represents a precarious situation that may be understood by appealing to attachment dynamics. Attachment anxiety, controlling for attachment avoidance, was remarkably highly associated with contempt-proneness in individuals who chose romantic partners as their contempt targets (see Table 14). Conversely, attachment avoidance, controlling for attachment avoidance, was nonsignificantly and slightly negatively associated with contempt-proneness in this same group, in contrast to the links observed between avoidance and contempt in the other groups and Study 4. This suggested that attachment anxiety, which involves hypervigilance toward threat, fear of rejection and separation, and perceiving others as unavailable and unresponsive to one’s needs, more heavily or consistently contributes to contemptuous feelings in this group and may explain why they chose their partners as contempt targets. Indeed, within the close and distal contempt target groups, the association between attachment anxiety and dispositional contempt increased the closer the nominated contempt target was to participants (see Table 14). Conversely, across groups, the association between attachment avoidance and dispositional contempt generally increased the further the target was to his or her nominator. This suggested that the need for non-interference and interpersonal space from others was more greatly associated with dispositional contempt in those who expressed wanting distance from those already distant to them.

Table 14.

Study 4: How Does Attachment Contribute to Dispositional Contempt In Individuals Who Nominated Romantic Partners As Contempt Targets?

Social Category	n	B		R2	F
		Anxiety	Avoidance
Close
Romantic partner	45	.68**	−.03	.45**	17.229
Friend	45	.53**	.27*	.51**	21.878
Immediate family	38	.23	.34*	.24**	5.376
Distal
Acquaintance	99	.23*	.44**	.57**	23.421
Co-worker	88	.39**	.20†	.51**	14.563
Extended family	63	.48**	.31**	.40**	19.958
Across Subsample	378	.40**	.28**	.32**	88.926

Note. Results are from multiple regression models in which attachment anxiety and avoidance were simultaneously entered as predictors of dispositional contempt for each group of participants who nominated a contempt target from the indicated social category.

^†p < .10;

^*p < .05;

^**p < .01

Study 5

So far, the preceding work, with the exception of Study 4, has largely consisted of correlational studies using self-report measures of personality. The purpose of Study 5 was to directly examine our conceptualization of contempt as a stable disposition that captures the tendency to react with contempt in relevant situations. We accomplished this in two ways: (1) by examining the temporal stability of dispositional contempt, and (2) by testing whether dispositional contempt predicted having contemptuous reactions to contempt-eliciting material. We had participants take the DCS twice separated by two weeks and also had them rate their reactions to six contempt-eliciting film clips and three film clips intended to elicit two other emotions – awe/admiration and pity/compassion – to try to ascertain the full range of emotional elicitors capable of evoking contempt in the contempt-prone.

Moreover, inspired by the controversy over whether people primarily feel contempt over violations of community (Rozin et al., 1997) vs. competence (Hutcherson & Gross, 2011), we wanted to know whether contempt situated at the level of personality was more sensitive to perceiving and reacting to targets demonstrating either type of violation – or both. Broadening our view of community vs. competence violations to encompass instances in which low warmth/affiliation vs. low competence/status are shown, we tested whether systematically manipulating the location of targets in interpersonal space (different combinations of low vs. high standing on these orthogonal dimensions) elicited different levels of contempt and, more importantly, whether these effects were moderated by one’s level of dispositional contempt. According to the stereotype content model (Fiske et al., 2002), different combinations of high vs. low warmth and competence evoke different emotional responses – contempt, pity, envy, or admiration – with contempt and similar emotions promoted by perceiving low competence and low warmth. Thus, contempt-prone people may be most sensitive to targets in this quadrant of interpersonal space.

Method

Participants and Procedure

Participants (Sample 11) were 182 MTurk workers (55.4% female) with a mean age of 40.7 (range = 20–72), and 81% were Caucasian. The DCS was administered at two time points separated by two weeks. Fifty-one percent of the sample returned for the second wave. Participants who did (M = 2.17, SD = .80) vs. did not (M = 2.01, SD = .71) return for the second wave did not differ in levels of dispositional contempt, t(180) = 1.45, ns. Across both waves, participants saw nine emotion-eliciting film clips (six contempt, three other emotion) in a randomized order. They rated their experience of various emotions, including “contempt” and “disdain,” at baseline (i.e., before watching the film clips) as well as with regard to their emotional experience immediately after having watched each video. It was then that the target of interest was identified, as we asked participants to rate the designated targets from film clips using a version of the DCS whose items were modified to reflect the elicitation of contempt as a state as opposed to tendency and the target from the film as the object of contempt. For example, after watching a video of a soldier making fun of a native boy outside his car, participants were asked to rate the extent to which they agreed with the statement that “Feeling disdain for the soldier came naturally to me.”

Measures

Emotional responses to films

Participants were given a list of 27 emotion terms and were asked to rate the degree to which each described their emotional experience on a 7-point Likert scale ranging from 0 (Not at all) to 6 (Extremely/a great deal) with a midpoint of 3 (Moderately). The other emotions were: admiration, amusement, anger, anxiety, awe, compassion, confusion, disgust, embarrassment, envy, fear, guilt, happiness, hate, interest, joy, love, pity, pride, resentment, sadness, shame, surprise, sympathy, and unhappiness. Ratings of contempt and disdain were averaged to derive what we refer to as “contempt label ratings.” Ratings of the DCS items that were modified to assess feelings of contempt for the identified targets of each film were averaged to derive what we refer to as “film DCS ratings.” Other measures from Study 5 were used in the foregoing studies. They included the Dispositional Contempt Scale, Buss-Perry Aggression Questionnaire, and Trait Pride Scale.

Films

Videos in the public domain (http://www.youtube.com) were rigorously evaluated and selected if they showed a clear example of a person violating a moral or competence standard or both. The location of targets in interpersonal space (different combinations of low vs. high standing on dimensions of warmth/affiliation vs. competence/status) was manipulated based on a-priori consensus among the authors (R.A.S., J.M.C., K.S.S.) regarding the attributes and actions that targets in the film clips conveyed. Making fun of others and being unfriendly are examples of low-affiliation behaviors/attributes, and showing physical disability or lack of intelligence are examples of low-power behaviors/attributes. An example of a target considered low on both dimensions was a young man speaking crudely about his physical relations with a woman while discussing – and misusing – a commonly known phrase (“vice versa”). We included a wide range of relevant behaviors/attributes, as it was possible that dispositional contempt would predict feelings of contempt for targets who were low on the vertical dimension (e.g., old; physically disabled) without being incompetent, per se. Film clips were shortened to last approximately 30–45 seconds. A description of film clip materials, including where targets were put in interpersonal space, can be found in Appendix 1.

Appendix 1.

Film Clip Scenarios for Study 4 in Decreasing Order of Mean Contempt Ratings (Contempt Label and Film DCS Ratings)

Quadrant	Film Clip	Contempt Rating		Top Three Other Emotion Ratings			r with DCS
		M	SD	Emotion	M	SD
Low W, High C	Soldiers are shown being cruel. One of them points and jeers at an unhealthy-looking boy native to the territory over which there is conflict.	3.64 3.80	2.08 .93	disgust	4.62	1.78	.08 .22**
				anger	4.09	2.06
				sadness	4.01	1.96
Low W, High C	A well-known radio show host makes fun of a celebrity who has Parkinson’s disease, mocking his symptoms and doubting their authenticity.	3.00 3.41	2.29 1.25	disgust	3.46	2.32	.20** .28**
				anger	3.10	2.30
				sadness	2.61	2.11
Low W, Low C	A young man, while lewdly discussing sexual acts, shows misunderstanding of the phrase “vice versa” but is accusing others of the same.	2.08 3.30	2.03 1.15	disgust	2.63	2.14	.23† .34**
				confusion	1.62	1.86
				embarrassment	1.52	1.87
Low W, High C	A male judge on a variety show openly rejects a candidate based solely on the candidate’s being African American. The judge is of Asian descent.	2.05 2.92	2.07 1.21	disgust	2.46	2.26	.16 .17
				surprise	2.49	2.11
				interest	2.19	2.00
Low W, Low C	A single mother living in a motel with her ten children blames the government for her situation and for not providing enough financial support.	1.86 2.94	1.82 1.12	sadness	2.66	2.06	.13 .35**
				disgust	2.61	2.12
				unhappiness	2.33	2.00
High W, Low C	A beauty pageant contestant who is asked a question that draws on world knowledge gives a response that suggests poor logic and ignorance.	1.10 2.24	1.59 0.96	embarrassment	2.98	2.14	.46** .56**
				amusement	2.86	2.23
				pity	2.56	2.15
High W, Low C	Elderly volunteers discuss their life-endangering mission to help clean the radiation caused by the Fukushima nuclear reactor crisis.	0.22 1.32	0.92 0.57	admiration	4.11	2.06	.24** .27**
				interest	3.49	1.87
				compassion	2.94	1.93
High W, Low C	A celebrity with Parkinson’s disease is trying to raise awareness of the disease in a public service announcement while he is showing symptoms.	0.18 1.28	0.71 0.56	compassion	4.48	1.85	.26** .28**
				admiration	3.63	2.18
				interest	3.50	2.02
High W, High C	A famous female comedian is giving a humorous speech in acceptance of a prestigious award for her talent.	0.11 1.37	0.46 0.56	amusement	3.48	2.27	−0.09 .14
				interest	3.08	1.94
				admiration	2.50	1.98

Note: W = warmth, C = competence; DCS = Dispositional Contempt Scale.

Results and Discussion

Is Dispositional Contempt Stable?

The stability coefficient of dispositional contempt was high (r = .86), suggesting that individuals’ levels of contempt-proneness were stable, at least over a two-week period. This supports the notion that dispositional contempt is an actual tendency that operates at the level of personality. It also suggests that this stable tendency is measurable with the DCS, providing further validation of our measure.

Does Dispositional Contempt Predict Experienced Contempt?

We first examined whether dispositional contempt predicted participants’ baseline ratings of contempt, as would be expected and be notable in its own right if contemptuous individuals, in everyday life, have a lower threshold for reacting with contempt, experience more elicitors, and have higher-intensity and/or longer-lasting episodes with little or no motivation or ability to regulate their reactions. Indeed, dispositional contempt was associated with baseline ratings of contempt (r = .44, p < .01) and disdain (r = .41, p < .01).⁴ We thus controlled for baseline ratings of contempt in the following analyses.

To examine whether contempt-proneness was differentially related to experiencing contempt for targets showing different combinations of low vs. high warmth vs. competence, we first ran correlations between individuals’ dispositional contempt scores and their mean contempt ratings for each quadrant. We found that dispositional contempt was most highly associated with experiencing contempt for targets who were in the High Warmth/Low Competence quadrant (r = .44, p < .01), followed by Low Warmth/High Competence (r = .24, p < .01) and Low Warmth/Low Competence (r = .21, p < .05); contempt-proneness did not predict contempt for High Warmth/High Competence targets (r = −.09, ns). Results held when controlling for anger- and hubristic pride-proneness (which had weaker or no associations with film-elicited contempt) and together suggested that dispositional contempt was most predictive of reacting with contempt when something about the target (e.g., smiling; helping behavior) signaled prosocial intentions all the while they demonstrated a lack of competence or power. Interestingly, in these cases, alternate – and indeed dominant – emotional responses were compassion, pity, and embarrassment, all conceptually opposite to contempt. Consistent with this interpretation, dispositional contempt was negatively associated with mean elicited compassion (r = −.24, p < .01) for targets in this and only this High Warmth/Low Competence quadrant. Dispositional contempt was unrelated to elicited pity and embarrassment, which showed positive relations with dispositional hubristic pride (rs = .18 and .25, respectively, both ps < .05); anger-proneness was not related to any of these “softer” emotional responses.

To better leverage our data to explore the effects of targets’ systematically varying in competence and warmth and whether these effects were moderated by participants’ levels of dispositional contempt, we performed a regression analysis using mixed effects modeling to control for repeated measurements. The dependent variable was film-elicited contempt and independent variables were high vs. low standing on warmth vs. competence (dummy-coded to yield two factors at two levels), dispositional contempt, and their two- and three-way interactions. Here, a significant interaction between warmth and competence would signify that targets’ standing on neither dimension alone explained contemptuous experience, which was better explained by their combination, i.e., targets’ location in a quadrant. Then, a significant interaction of either dimension (or their combination) with dispositional contempt would signify that dispositional contempt moderated the degree to which targets’ standing on these dimensions elicited contempt.

We observed main effects for warmth (b = −1.30, SE = 0.31), competence (b = .75, SE = 0.26), and dispositional contempt (b = .50, SE = 0.11), all ps < .001, and a marginally significant two-way interaction between competence and dispositional contempt (b = −.19, SE = 0.11, p < .10). With the three-way interaction non-significant, we dropped it from the model and reevaluated effects. Here, we found main effects of warmth (b = −1.12, SE = 0.24), competence (b = .94, SE = 0.19), dispositional contempt (b = .56, SE = 0.10), and two two-way interactions between competence and warmth (b = −.67, SE = 0.11) and competence and dispositional contempt (b = −.27, SE = 0.08), all ps < .001.

We decomposed the two-way interactions that competence had with warmth and dispositional contempt, respectively. We found that warmth was negatively associated with film-elicited contempt regardless of levels of competence but that competence was negatively associated with elicited contempt when warmth was high (b = −.27, SE = 0.06, p < .001) and positively when warmth was low (b = 0.34, SE = 0.10, p < .01). This is consistent with accounts of moral contempt inasmuch as showing a disregard for others coupled with actual ability to negatively affect them is represented by the Low Warmth/High Competence quadrant. Note that this result speaks to elicited contempt across the sample, that is, irrespective of dispositional contempt. However, the interactive effect of competence with dispositional contempt indicated that dispositional contempt predicted more film-elicited contempt when competence was low (b = .51, SE = 0.07, p < .001) and not when it was high (b = 0.17, SE = 0.10, ns), as shown in Figure 2. These results suggest that contempt situated at the level of personality is more sensitive to perceiving and reacting to low competence/power behavior, consistent with Hutcherson and Gross’ (2011) account (see also Ufkes et al., 2011).

Figure 2.

Plot of simple slopes showing the interaction between participants’ level of dispositional contempt and film-targets’ level of competence (high vs. low) in the prediction of elicited contempt as measured with a version of the DCS whose items were modified to assess feelings of contempt for the targets of each film.

Study 6

In Study 5, we demonstrated that dispositional contempt is trait-like in its stability and that it predicted state-related changes in emotional responding to contempt-eliciting material. In Study 6, we examined whether dispositional contempt predicted interpersonal responding in the realm of romantic relationships, which critically define one’s social environment (Roberts, Robins, Caspi, & Trzesniewski, 2003). Seminal work by Gottman (1994) identified contempt as the most pernicious threat to one’s romantic ties. We asked not only if one’s level of dispositional contempt, over and above Disagreeableness, impacted relationship outcomes, but also if perceiving one’s partner as contemptuous had ill effects. Both would be expected if the social role of contempt is to distance and reject others who violate standards (Fischer & Roseman, 2007). For one, contempt-prone individuals would be more likely to find faults in their partners, disdain them, want nothing to do with them, and sever the relationship. However, another interesting idea is that because contemptuous behaviors are evolutionarily designed to push others away and thus may act as a “repellant,” perceiving a partner as contemptuous may, in a very basic sense, make them less attractive. This would further be the case if a partner was viewed as contemptuous because one was regularly a recipient of his or her contempt, resulting in drops in self-esteem and lowered fondness (Melwani & Barsade, 2011). With these considerations in mind, we pitted these sources of contempt against each other to determine if poor relationship outcomes were primarily explained by one’s own dispositional contempt or that perceived in a partner. We had participants come into the lab at the beginning and end of a month, at which points they reported on their own and their partners’ levels of dispositional contempt and Disagreeableness as well as their relationship functioning.

To examine effects on relationship functioning, we turned to the investment model (Rusbult, 1980), according to which the most important predictor of relationship success is commitment, or the sense of and desire to maintain attachment to one’s partner (Rusbult, Johnson, & Morrow, 1986). Three factors are posited to contribute to commitment (Rusbult & Buunk, 1993). We predicted that contempt-proneness would be associated with low satisfaction, which reflects the extent to which the relationship is meeting or exceeding one’s standards (Rusbult, 1983). In addition, given the contemptuous person’s tendency to make social comparisons, we predicted that contempt-proneness would be related to perceiving a greater quality of alternatives, which is fostered by looking for and attending to them. Finally, because competing hypotheses could be generated in light of the resource-saving function of contempt, we had no specific predictions for how contempt-proneness would be associated with investment, or feeling much at stake in the relationship due to investments lost if it were to end.

Method

Participants and Procedure

Participants (Sample 12) were 165 undergraduate students (81% female) enrolled in a psychology course who participated for course credit. Participants had a mean age of 19.8 (range = 18–37) and were ethnically representative of the university population (32% Asian, 27% Caucasian, 18% Latino, 2% African American, 4% Middle Eastern, and 17% other ethnicity). Of these participants, 95.7% were in romantic relationships. 90.7% percent of participants described themselves as heterosexual; 5.6%, homosexual; 2.5%, bisexual; and 1.2%, as other. Measures, including the DCS, were administered at two time points separated by one month. Eighty-five percent of the sample returned for the second wave, and of the participants who were in relationships at Time 1 who returned for Time 2, only 9, or 6.7%, experienced a break-up in the interim.

Measures

Dispositional Contempt and Disagreeableness

As in our previous studies, the 10-item DCS was used to assess level of dispositional contempt. The nine-item agreeableness subscale of the Big Five Inventory was used to assess Disagreeableness.

Relationship Functioning

We used the Investment Model Scale (IMS; Rusbult, Martz, & Agnew, 1998) to assess relationship commitment, satisfaction, quality of alternatives, and investment. Participants used a 7-point Likert scale to rate their agreement or disagreement with 25 statements (e.g., “I am committed to maintaining my relationship with my partner,” “I feel satisfied with our relationship,” “My needs for companionship (doing things together, enjoying each other’s company, etc.) could be fulfilled in alternative relationships,” “I have invested a great deal of time in our relationship”). Reliability coefficients for the four subscales were in the mid-.80s based on data collected at the first time point.

Results and Discussion

Is Dispositional Contempt Stable?

The stability coefficient of participants’ levels of dispositional contempt was high (r = .75), suggesting that, over a one-month period, participants’ levels of contempt-proneness were trait-like. With more data available at two time points than in Study 5 and with the time points’ spaced further apart, we more thoroughly assessed the stability of dispositional contempt by examining its construct stability (i.e., factorial invariance). We fit a 1-factor model with no constraints specified beyond fixing the first factor loading to one for factor identification (i.e., configural invariance). This first model demonstrated satisfactory fit (χ²(163, N = 161) = 281.55, p < .01; CFI = .91; TLI = .90; RMSEA = .07, 90% CI [.05, .08]). We next fit the same 1-factor solution with the modification that corresponding factor loadings between occasions were constrained to be equivalent (i.e., weak configural invariance). This nested model also showed satisfactory fit (χ²(172, N = 161) = 287.31, p < .01; CFI = .91; TLI = .91; RMSEA = .07, 90% CI [.05, .08]). Chi-square difference tests indicated that dispositional contempt (Δχ²(9) = 5.76, p = .76) demonstrated weak invariance between measurement occasions, suggesting that DCS items were contributing the same information about dispositional contempt at each time point. We replicated analyses for perceptions of partners’ dispositional contempt, and results again indicated weak invariance between measurement occasions.⁴

Assessed this way, test-retest correlations were large for participants’ own levels of dispositional contempt (r = .84) and perceptions of partners’ dispositional contempt (r = .89) (self- and other-ratings were themselves similarly related at both time points, rs = .34 and .41 at Times 1 and 2). Test-retest correlations were also high for participants’ own levels of Disagreeableness (r = .91) and partners’ Disagreeableness (r = .95) (with self- and other-ratings similarly related to each other at both time points, rs = .41 and .36 at Times 1 and 2). Consistent with Study 5, these stability coefficients supported the notion that dispositional contempt is a stable individual difference variable with a consistent psychological structure as per the DCS. These results also suggested sufficient between-observation stability in the factor structure of dispositional contempt to justify analyses at the latent construct level. Factor scores were saved for use in hierarchical and backward stepwise regression analyses to identify the personality tendencies that predicted these relationship variables, and structural equation modeling was used to analyze lagged relationships between them.

Is Dispositional Contempt Uniquely Associated with Relationship Functioning?

We first tested the ability of dispositional contempt, over and above Disagreeableness, to predict relationship variables at Time 1 using hierarchical regression analysis. With the hypothesis that dispositional contempt would be a unique predictor of these variables, we used the strategy of entering one’s own and partner’s perceived dispositional contempt in the first block. If the predictive ability of this block was significant at p < .05, it was retained; otherwise, it was dropped from the model. In the second block, we included one’s own and partner’s perceived Disagreeableness. Again, if this block was significant, it was retained; otherwise it was dropped.

Commitment

For commitment, there was a significant effect of the dispositional contempt block (F= 5.91, p < .01, R² = .13) and no significant change in the model when the Disagreeableness block was entered (F = 0.37, p = .69, ΔR² = .008). This resulted in a final model with one’s own (B = −.04, SE = .14, p = .75) and partner’s perceived (B = −.55, SE = .19, p < .01) dispositional contempt predicting commitment and suggested that the effect of this block was driven by partner’s perceived dispositional contempt.

Satisfaction

For satisfaction, there was a significant effect of the dispositional contempt block (F = 9.61, p < .01, R² = .19) and no significant change in the model when the Disagreeableness block was entered (F = 0.64, p = .53, ΔR² = .013). This resulted in a final model with one’s own (B = −.22, SE = .21, p = .29) and partner’s perceived (B = −.70, SE = .16, p < .01) dispositional contempt predicting satisfaction, and again suggested that the effect was driven by partner’s perceived dispositional contempt.

Quality of Alternatives

For alternatives, there was a nonsignificant trend of the dispositional contempt block (F = 2.33, p = .10, R² = .06), so it was dropped. There was also a nonsignificant effect of the Disagreeableness block (F = 0.53, p = .59, R² = .013). This suggested that neither the contempt nor Disagreeableness blocks consisted of strong predictors of alternatives. This was surprising given a significant correlation between alternatives and self-rated dispositional contempt (r = .25, p < .05) but suggested that common variance between the dispositional contempt predictors diminished the association between alternatives and self-rated dispositional contempt. This suggested that the only disposition likely to be related to alternatives was one’s own level of dispositional contempt.

Investment

For investment, there was a nonsignificant trend of the dispositional contempt block (F = −1.45, p = .24, R² = .04), so it was dropped, but there was a significant effect of the Disagreeableness block (F = −3.44, p = .04, R² = .08). This resulted in a final model with both one’s own (B = −.66, SE = .36, p = .07) and partner’s perceived (B = −.28, SE = .31, p = .36) Disagreeableness predicting investment, and suggested that the Disagreeableness block effect was driven by self-rated Disagreeableness. As with alternatives, common variance between the Disagreeableness predictors may have reduced the unique variance attributable to self-rated Disagreeableness.

Replicating the Models

To see if we would arrive at the same models identified with our hierarchical approach, we conducted a backward stepwise regression that iteratively selected the optimum predictors for the regression model from a full model by maximizing fit based on AIC. The optimum models identified for each outcome were as follows: (1) one’s commitment was best predicted by partner’s perceived dispositional contempt (B = −.58, SE = .17, p < .01, R² = .13); (2) one’s satisfaction was best predicted by partner’s perceived dispositional contempt (B = −.61, SE = .14, p < .01, R² = .18); (3) quality of alternatives was best predicted by one’s own dispositional contempt (B = .46, SE = .22, p = .03, R² = .05); and (4) investment was best predicted by one’s own Disagreeableness (B = −.80, SE = .33, p = .02, R² = .07). Overall, this data-driven approach largely replicated our results from the hierarchical regression procedure.

Finally, we wanted to see if we would replicate our models for Time 1 in Time 2. To do this, we treated Times 1 and 2 as time groups, and each outcome was modeled using the foregoing Time 1 models identified with our hierarchical and stepwise procedures with a constraint made such that Time 2 effects were equivalent with Time 1 effects. There was a significant effect of partner’s perceived dispositional contempt for commitment (B = −.51, SE = .10, p < .01), with excellent model fit. Similarly, there was a significant effect of partner’s perceived dispositional contempt for satisfaction (B = −.62, SE = .09, p < .01), with excellent model fit. Alternatives had a significant effect of one’s own dispositional contempt (B = .45, SE = .15, p < .01), with excellent model fit. Finally, investment had a significant effect of participants’ own Disagreeableness (B = .78, SE = .23, p < .01), with excellent model fit. Taken together, results suggested stability in the relation of dispositional contempt vs. Disagreeableness with respect to the relationship variables of interest, indicating that their predictive utility was robust across these observations.

Does Dispositional Contempt Prospectively Predict Relationship Functioning?

Due to the lagged nature of the Time 1 and 2 observations, we completed our analysis by testing for cross-lagged relations between the relationship variables and their optimum predictors as per our previous analyses. For each model, Time 1 and 2 measurements of outcomes and predictors were freed to allow estimation of unique covariance. Time 2 outcomes were regressed on Time 1 outcomes and predictors. Time 2 predictors were regressed on Time 1 outcomes and predictors. For identification of the model, a nonsignificant cross-loading was identified and then constrained to zero; fit for all models was excellent.

Two consistent patterns were identified across these cross-lagged models: (1) correlations between the outcome and predictor were larger at Time 1 than 2, and (2) the autoregressive effects of all variables at Times 1 and 2 were so high that cross-loadings between predictor and outcomes (e.g., Time 1 perceptions of partner’s dispositional contempt on Time 2 commitment) were not significant. That is, the common variance between outcomes and predictors was better explained by their previous observations such that little residual variance from outcomes and predictors at different time points could be related. Thus, there were no cross-lagged effects of the relationship variables and their identified predictors.

Taken together, results from Study 6 established dispositional contempt – one’s own and to an even greater extent that perceived in one’s partner – as uniquely predictive of three out of four variables related to romantic relationship functioning. This included the one – commitment – that has been established in the literature as most central for predicting a relationship’s success (Le & Agnew, 2003). Thus, findings suggest that perceiving a partner to be contemptuous has a more negative effect on one’s own commitment and satisfaction than being contemptuous oneself. Although surprising, this asymmetry makes sense. The tendency to feel contempt for others may not directly translate to feeling contempt for one’s partner; indeed, a close relationship partner who has effectively made the cut amidst so many perceived good-for-nothings may be a safe haven to the contemptuous person. In contrast, perceiving a partner to be contemptuous may be more uniform in its effects; it implies a partner does not appear to care about or get along with others, that they may be a threat to one’s own self-esteem and sense of belonging, and that they might just be unpleasant to be around. Either way, results suggest that contempt-proneness is unattractive. Future work might examine whether partners are accurately perceived as contemptuous and if partners’ commitment is influenced by one’s own dispositional contempt, even if one does not realize it and even if one’s own are not

General Discussion

The present research provides the first empirical look at the contemptuous person, who has a tendency to view others as falling short of his or her standards and to look down on them with a tendency to distance and/or derogate them. Despite the pervasiveness of contempt and its candidacy as a “basic” emotion (Ekman, 1994), the existence and nature of stable individual differences in experiencing and expressing this emotion have not been addressed in previous research. To enable an exploration of dispositional contempt, we first developed the Dispositional Contempt Scale (DCS), a 10-item self-report measure with excellent psychometric properties (coefficient alphas ~.90s; test-retest reliability = .85 or more over several weeks) and a unidimensional structure whose emergence at every stage of scale development, during which we sampled content for different standard violations (e.g., competence, morality), suggested that contempt-proneness is a general tendency not driven by detecting shortcomings in any particular domain. Associations of the DCS with theoretically related constructs were large enough to support our predictions but not so large so as to suggest that contempt-proneness is redundant with them. Using the DCS, we demonstrated that dispositional contempt has a range of elucidating correlates and important real-world implications.

Figure 3 presents a schematic of how results from our analyses, which largely confirmed our hypotheses, might be integrated into a theoretical account of dispositional contempt that maps onto the cognitive, subjective, and behavioral features of contempt as an emotional reaction. As shown in Figure 3, its antecedent appraisals (e.g., “Someone is violating my standards”) directly guided our reasoning as to factors that may confer vulnerability to experiencing contempt (e.g., chronic attunement to standards violations, as seen in perfectionism). Considering its subjective feeling (e.g., unpleasant; social aversion; inflated self) helped identify constructs likely to be in its nomological network and psychological space (e.g., Disagreeableness; hubristic pride). And the distancing and rejecting behaviors that accompany contempt, which couple tightly with its social functions (Fischer & Roseman, 2007), motivated ideas about the toxic interpersonal dynamics created and poor interpersonal outcomes met by the contempt-prone.

Figure 3.

Conceptual model depicting dispositional contempt and its associated constructs, including how these associations map onto what has been established about contempt as an emotion.

Dispositional contempt was both related to but distinct from theoretically similar emotion dispositions in Study 1. It was most highly associated with dispositional envy, anger, and hubristic pride. The high correlation with dispositional envy, a negative response to others’ succeeding at meeting important standards, suggested that contempt-prone individuals are sensitive to social evaluation and their standing on the status hierarchy. Similar links were observed with hubristic pride, with similar conclusions drawn. Unexpectedly, although disgust is a theoretical cousin to contempt and anger (Rozin et al., 1999), dispositional contempt was mostly unrelated to dispositional disgust, including dispositional moral disgust, as measured with the Three-Domain Disgust Scale (Tybur, Lieberman, & Griskevicius, 2009). It is advisable to use other measures of disgust sensitivity before definitively claiming that dispositional contempt is unrelated to it.

Importantly, we suspect that contemptuousness is inversely related to a host of positive emotion dispositions (Shiota, Keltner, & John, 2006). Future work should examine the nature and strength of these associations to determine which positive emotion dispositions contempt-proneness is particularly unlike and thus which emotions are most likely to be contempt’s antithesis. Gottman (1994) wrote that “like most couples I’ve worked with over the years, [they] wanted just two things from their marriage – love and respect” (p. 18). He further stated that respect was the “opposite of – and antidote for” (Gottman, 1994, p. 61) contempt. A logical next step, then, is to examine the association between dispositional contempt and emotional tendencies grounded in having respect for and/or affection for others, such as being prone to admiration, awe, love, and compassion. This would allow us to further delineate the nomological network of contempt and better situate contempt within the extant psychological literature.

In terms of the broader dimensions of the Big Five, dispositional contempt was most strongly related to Disagreeableness with correlations in the .70s. Although this association might seem condemning for recognizing dispositional contempt as unique, results from factor analyses suggested that contempt-proneness and Agreeableness are statistically distinct, and we ultimately treated their relation as informative.

One possibility is that contempt-proneness is a facet defining the low pole of Agreeableness. Whereas Agreeableness provides high “bandwidth” for conveying the overall range of a person’s prosocial tendencies, dispositional contempt could offer greater “fidelity” for predicting, describing, and explaining people’s specific reactions to others’ missteps and the impact of these reactions on their social relations. Consistent with this gain in precision, we found that contempt-proneness had incremental validity over Disagreeableness in its association with variables related to personal standards, social status, and relationship functioning. Although consensus has yet to be reached about the precise facets of Agreeableness or any other Big Five domain, we recommend that the DCS be examined alongside extant facet-level measures of the Big Five, such as the NEO Personality Inventory–Revised (McCrae & Costa, 2008), Big Five Aspects Scale (DeYoung et al., 2007), and the BFI-2 (Soto & John, in press). It would be helpful to know, for example, if, of the BFI-2 Agreeableness facets, contempt-proneness is more negatively related to Respectfulness, Trust, or the more factor-pure Compassion. Another possibility is that contempt is the “affective core” of Disagreeableness (see Figure 3). Identifying contempt as part of the affective underpinnings of Disagreeableness could illuminate the mechanisms by which Disagreeableness is sustained and brings about important life outcomes.

Dispositional contempt was related to each of the constituents of the Dark Tetrad – narcissism, psychopathy, Machiavellianism, and sadism – in Study 2. This is consistent with the social-distancing and status-differentiating roles of contempt and the links that dispositional contempt has with anger, hostility, and low guilt-proneness. Of the Dark Tetrad traits, dispositional contempt was most strongly related to Machiavellianism, more so than was hubristic pride, suggesting that contempt-proneness involves a greater preoccupation with status and willingness to manipulate others to attain it. Interestingly, “contemptuous delight” (Bursten, 1972) has long been noted as a motivator for exploiting others. Researchers and lay people alike devote a great deal of effort toward trying to understand the dark personality tendencies and other processes that guide antisocial behavior. Our research suggests that contempt-proneness offers an important new lens for doing so. The DCS could be administered to prison populations to determine if contempt-proneness is uniquely associated with criminality, including violent criminality, as Tausch et al.’s (2011) research on contempt in intergroup contexts suggests.

Although associated with callousness, contempt-proneness was related to having a fragile self. It was moderately related to overt narcissism and highly related to covert narcissism (Hendin & Cheek, 1997), for which a devaluing facet has been identified in previous research (Pincus et al., 2009). Dispositional contempt was related to having low self-esteem. It could be that contempt-prone individuals are so mired in monitoring standards and differentiating status that their self-views suffer and they demean others to regain equilibrium. We tentatively view perfectionism and low self-esteem as predisposing factors, as indicated in our model, and sssessing perfectionistic tendencies, self-esteem, and contempt at several time points would allow researchers to capture their causal dynamics (Figure 3). Moreover, while we assessed only level of self-esteem, more nuanced aspects, such as its instability, may be part of the unique signature of dispositional contempt.

Studies 3 and 4 showed that contempt-prone individuals were low in attachment security, with associations found between contempt-proneness and attachment avoidance and anxiety. These findings add to literature mapping the functional relations between attachment styles and discrete emotions (Mikulincer & Shaver, 2005; Simpson et al., 2007). Given the stability of attachment over the life course (Fraley, 2002), we speculate that the caregiving experiences and temperamental factors associated with insecure attachment contribute to contempt-proneness. A history of perceiving caretakers as unavailable in times of need, for example, may predispose one to seeing others as generally unreliable in their roles and responsibilities. Developmental work that tracks the emergence and course of contempt-proneness across the lifespan would thus be invaluable. We also see attachment as helpful for informing what the contemptuous person is not. Securely attached individuals show an affective style that is open and adaptable with access to a wide range of emotional responses (Cassidy, 1994; Malatesta, 1990; Magai, Hunziker, Mesias, & Culver, 2000). This flexibility is not expected in contempt-prone people.

Dispositional contempt predicted contemptuous reactions to eliciting material in Study 5. Contempt-proneness predicted reacting with “contempt” and “disdain” and providing high ratings on the state version of the DCS upon exposure, through film clips, of others’ demonstrating standard violations. Although mean contempt ratings across our sample were highest for moral violations, dispositional contempt was most predictive of individual differences in contempt elicited by incompetence. This difference may have arisen from there having been a set of diametrically opposed but appropriate alternate emotional responses – namely, pity, compassion, and embarrassment – whenever inept or otherwise weak individuals were also well-intentioned. Indeed, dispositional contempt was negatively associated with reacting with compassion. Consistent with this set of results, Hutcherson and Gross (2011) found that pity was the primary emotion activated by their incompetence vignettes, more so than contempt or the other emotions they examined, anger and disgust (see also Fiske et al., 2002). Also, although Study 5 showed that contempt-proneness predicted self-reported contemptuous experience, future work should examine whether it predicts contemptuous reactions that are overtly expressed and observable to others. This would be a critical next step.

Finally, we demonstrated in Study 6 that dispositional contempt is a unique predictor of relationship functioning. Although one’s own contemptuousness was expected to be toxic for one’s relationship, results suggested that seeing one’s partner as contemptuous was more harmful, being related to less commitment and satisfaction. If staying in a relationship is a function of the degree to which attractors overwhelm repellants in the relationship (Adams & Jones, 1997; Arriaga & Agnew, 2001; Rusbult & Buunk, 1993), it may be that contempt, evolutionarily designed to distance others, is a powerful repellant. Indeed, past work has shown that relationships are supported by feeling idealized (Murray, Holmes, & Griffin, 2000) or truly known and accepted (Swann, De La Ronde, & Hixon, 1994) by partners, for whom warmth and kindness are cited as desirable qualities (Sprecher & Regan, 2002). None of the above are likely with partners who are perceived as contemptuous. Future work might examine the extent to which these perceptions were accurate and whether they came from having been a target of contempt vs. an observer of contempt directed at others. We had asked only about the perception of a partner’s contempt for others in general, but the DCS can be modified to assess contempt for specific targets, including oneself as a target of a partner’s contempt.

Our work was not without limitation. As three of our six studies were correlational and cross-sectional, our research cannot speak to the dynamic interrelations and causal ordering among dispositional contempt and other variables (see Figure 3). Experimental or longitudinal work is needed to map out their temporal relations and pinpoint mediating mechanisms. Even in Study 6, the high stability of our variables measured at the beginning and end of a month made it difficult to detect interplay among them. If contempt is facilitated by interpersonal distance (Fischer & Roseman, 2007), or if having a bad relationship could cast a general pall over one’s views of others, lower commitment and satisfaction might impact one’s contempt-proneness. More intensive time-series sampling over shorter intervals could be used to reveal lagged relations between these outcomes and predictors (e.g., Castro-Schilo & Ferrer, 2013; Conroy, Ram, Pincus, & Rebar, 2015; Gonzales & Ferrer, 2014). Such methods would also shed light on how day-to-day changes in contempt are linked to individual differences in contempt-proneness more generally. In addition, future work would profit immensely from using more than paper-and-pencil measures and, when using these measures, multiple sets of reporters.

With dispositional contempt established as a relatively distinct construct able to uniquely predict outcomes, many questions to address in future research emerge. For example, although we speculated to some extent, why are some people more contemptuous than others? What accounts for the individual differences? Answers may be rooted in genetic, hormonal, cognitive, and environmental factors. Also, with the consequences of dispositional contempt better understood, we are well-advised to explore how to reduce it. Is contempt-proneness a malleable tendency that can be changed over time by focusing on others’ laudable qualities or status as human beings of intrinsic worth despite fallibility? Loving-kindness meditation (Hutcherson, Seppala, & Gross, 2008) and compassion training (Fredrickson et al., 2008, Klimecki et al., 2014) may be helpful in this regard. Future work should also explore whether amplifying authentic self-love reduces contempt in lasting ways. Processes that decrease contempt may, according to our research, promote mental and behavioral flexibility; boost self-esteem; expand one’s social network; lower loneliness and depression; cement romantic relationships; and overall generate more caring members of humanity. Thus, dispositional contempt offers a new theoretical and empirical framework for understanding an emotion that, while understudied, critically affects how people relate to each other and, likely, ultimately to themselves.