Abstract
Current research into the elicitors of nonconscious behavioral mimicry focuses almost exclusively on prosocial motives. Alternatively, the present research investigates whether the self-centered desire to be liked by others also induces mimicry. We investigate this issue by measuring the mimicry behavior of narcissists—a “dark personality” that is uniquely characterized by a desire to be liked by rather than by genuine liking for people. Narcissists are particularly motivated to form social alliances with high-status others. Hence, it was hypothesized that narcissistic participants would show more mimicry of higher status others (compared to lower status others). Support for this hypothesis was provided by a mixed within- and between-subjects field experiment in which the mimicry behavior of participants identified as being above or below average in subclinical narcissism was observed in a scripted interaction with a higher and lower status other.
When meeting someone for the first time, on what basis do we judge whether they are a friend or a foe? One of the subtle nonverbal cues that influences our perception of a person’s social motives is their behavioral mimicry, or the extent to which they nonconsciously imitate our bodily, vocal, or facial expressions (for a review, see Chartrand & van Baaren, 2009). When a person subtly mimics their interaction partner, they are perceived to be more likable (Chartrand & Bargh, 1999), empathic (Stel & Vonk, 2010), and trustworthy (Maddux, Mullen, & Galinsky, 2008). To the extent that someone mimics us, they are perceived to be a friend rather than a foe. But is this necessarily the case?
Current research into the elicitors of nonconscious behavioral mimicry focuses almost exclusively on communal or prosocial motives. For example, despite being potentiated by an automatic ideo-motor process (Chartrand & Bargh, 1999), the expression of behavioral mimicry is facilitated by the motivations to affiliate (Lakin & Chartrand, 2003; Stel & Vonk, 2010), and perceptions of closeness and interdependence (Ashton-James, van Baaren, Chartrand, Decety, & Karremans, 2007). Alternatively, the present research investigates whether mimicry may be facilitated by the self-centered desire to be liked by others. We investigate this issue by measuring the mimicry behavior of narcissists—a personality type whose behavior is characterized by a desire to be liked by rather than by genuine liking for people (Campbell, 1999; Emmons, 1987).
The Desire to be Liked as an Elicitor of Mimicry
The desire to be liked is of particular interest to us because it is associated with a range of often automatic affiliation behaviors, but is independent from liking. For example, waiters increase their chances of being liked by customers by raising their eyebrows, heightening their vocal pitch, seeking eye contact, and shifting their posture forward. Much of this nonverbal behavior is enacted nonconsciously, not in response to a genuine liking for all customers, but rather in response to a desire to be liked by customers (for larger tips and positive word of mouth; Grandey, 2000). As such, while research to date has measured mimicry only in response to a prosocial orientation toward others, it is possible that mimicry is also elicited in response to a desire to be liked that is not necessarily linked to prosocial motives.
Supporting the notion that mimicry may be expressed in response to the desire to be liked by others, Lakin, Chartrand, and Arkin (2008, experiment 2) demonstrated that after being excluded by an in-group member (in this case, someone of the same gender), participants showed greater mimicry of other in-group members compared to out-group members. Additional analyses revealed that participants’ mimicry of in-group members following their exclusion was predicted by the degree to which it was important for them to belong to their in-group. While the authors did not measure liking for or desire to be liked by the interaction partner, we interpret participants’ increased mimicry as a response to their heightened desire to be liked by in-group members, whose support is an important source of self-esteem and personal security (Williams, 2007), rather than a response to liking for their interaction partner, by whom they were virtually just excluded.
The Desire to be Liked and Narcissism
Subclinical narcissism is a personality typified by the desire to be liked in the absence of genuine liking. One of the three “dark” or antisocial personalities identified by Paulhus and Williams (2002), narcissists are at best indifferent toward other people in general (Emmons, 1987) and are at worst characterized by arrogance, dominance, extreme self-centeredness, manipulation of others for self-promotion, and a lack of empathic concern for others (Colvin, Block, & Funder, 1995; Emmons, 1984; Morf & Rhodewalt, 2001; Raskin & Terry, 1988). Like the other two dark personalities (Machiavellians and psychopaths), narcissists harbor particularly hostile feelings toward people whose success, intelligence, or popularity cannot be used for their own personal gain and threatens their feelings of personal superiority (Morf & Rhodewalt, 2001). Unlike Machiavellians and psychopaths, however, narcissists have a desire to be liked—by those whom they perceive to be inferior to them and by higher status others in particular.
For the narcissist, social interactions are not opportunities for social connection, but “settings for the enactment of social manipulations and self-presentations designed to engineer positive feedback or blunt negative feedback about the self” (Morf & Rhodewalt, 2001, p. 181). As such, narcissists are often more popular at first acquaintance than non-narcissists, being described as charismatic, attractive, competent, affable, and good-humored (Back, Schmuckle, & Egloff, 2010; Paulhus, 1998; Young & Pinsky, 2006). Narcissists are particularly charming toward others who are successful, intelligent, attractive, or popular, seeking to sustain their inflated self-image through strategic social alliances with other individuals who are deemed to be superior in some way (Campbell, 1999; Morf & Rhodewalt, 2001).
Narcissists are, therefore, like wolves in sheep’s clothing; their chronic desire to be liked by others motivates them to hide their feelings of indifference or even hostility toward others under a cloak of charm, friendliness, and interpersonal warmth (Back et al., 2010; Carlson, Vazire, & Oltmanns, 2011; Paulhus, 1998). Thus, by observing the nonconscious mimicry behavior of extreme narcissists who have a chronic goal to be liked by others—particularly others with higher social status than themselves—the present research aims to uncover whether mimicry can be facilitated by a desire to be liked.
Hypotheses
In general, people exhibit mimicry of each other’s behavioral gestures (i.e., foot shaking, head nodding, face touching) during an initial interaction (Chartrand & Bargh, 1999). However, mimicry appears to be undercut by interpersonal dominance and hostility toward others (McHugo, Lanzetta, Sullivan, Masters, & Englis, 1985). Since narcissism is a “dark personality,” associated with a lack of genuine interest in or liking for others, one might expect narcissistic individuals to mimic others less than non-narcissistic individuals. However, unlike the “dark personalities” of Machiavellianism and psychopathy, that are antisocial toward others in general and openly aggressive toward authorities or higher status others (Paulhus & Williams, 2002), narcissism is associated with the motivation to form social alliances with high-status others, who may be successful, intelligent, popular, or attractive (Campbell, 1999). If, as argued in the present research, mimicry behavior may be driven by a desire to be liked, then individuals who are high in narcissism should show greater mimicry of higher status others compared to lower status others, in the absence of greater liking for these individuals. The mimicry behavior of non-narcissists, on the other hand, should be less sensitive to interaction partners’ social status.
Method
Design and Participants
The forgoing hypotheses were tested in a mixed within- and between-subjects field experiment. The mimicry behavior of 91 participants (51 female, 40 male) was observed during two separate interactions with a higher and a lower status other before they completed a confidential survey.
Procedure and Materials
A research assistant playing the role of an “undergraduate assistant” to a “PhD researcher” recruited participants in university cafeterias. Participants were informed that the purpose of the study was to investigate interpersonal behavior and personality, and would involve two short interviews about university life (e.g., “describe a typical Friday from the moment you wake up in the morning to the moment you go to sleep”) and a questionnaire. In return for volunteering, participants received a chocolate bar. After providing consent to participate, the participants were interviewed—first by the “PhD researcher” and then by the “undergraduate assistant.” Participants' mimicry behavior was observed and recorded during both interviews. After the interview, participants completed a confidential survey containing measures of narcissism, a manipulation check, and a demographic survey.
Status Manipulation
Two research assistants (1 male, 1 female) were selected to conduct the experiment based on their demonstrated ability to role-play a PhD researcher (which stereotypically has higher intelligence than the average undergraduate participant) and an undergraduate research assistant. The male research assistant was assigned to the role of “PhD researcher” due to his age (slightly older than the average undergraduate student), height (a signal of status; Roberts & Herman, 1986), and ease with which he was able to make use of “power poses” (i.e., expansive, with open limbs; Carney, Cuddy, & Yap, 2010) during their interview with the participant, and because males are stereotypically higher in social status than females. The female research assistant was assigned to the role of “undergraduate assistant” because she was younger, had a higher pitched and softer voice (both submissive cues; Ohala, 1994), and was comfortable interacting in submissive postures (i.e., constricted, with legs together; Argyle, 1988).
Mimicry
During each interview, the interviewing researcher enacted 10 scripted behaviours (e.g., hair touch, posture change, leg cross, arm fold, nose touch), one per question, while the `observing' researcher recorded every time the participant mimicked scripted (but not any unscripted) gestures enacted by the participant. The researchers were trained to refrain from bodily movements that were not scripted (e.g., head nodding) as much as possible to minimize experimenter effects on the mimicry behaviour of participants. For a subset of participants, researchers recorded the extent to which they were mimicked during the interview in order to verify the accuracy of observer coding. For this sub- set of participants, the reliability of reported mimicry of the experimenter (r(52) = .79, p < .001) and assistant (r(52) = .41, p = .002) was good.
Narcissism
The desire to be liked, particularly by high-status others, is a unique characteristic of narcissism. For the purposes of identifying people who are high in narcissism, participants completed the Short ‘Dark Triad’ inventory (SD3; Jones & Paulhus, 2012), which includes 9 items measuring narcissism (e.g., “I hate being at the centre of attention (reversed)”; “I like to get acquainted with important people”), 10 items measuring Machiavellianism (e.g., “Most people are suckers”; “Make sure your plans benefit you, not others”), and 9 items measuring psychopathy (e.g., “I like to get revenge on authorities”; “I like to pick on losers”). The narcissism subscale of the SD3 has shown good reliability in previous studies conducted in the United States (Baughman, Dearing, Giammarco, & Vernon, 2012, and Lee, Ashton, Wiltshire, Bourdage, Viser, & Galluci, in press).
Manipulation Check
The post-interview survey also asked participants to use a 7-point scale (1 = low or not at all, 7 = high or extremely) to indicate their impressions of the PhD researcher and undergraduate assistant (separately) on each of the following dimensions: likeable, attractive (control variables), and as a manipulation check, intelligent. Single-item measures were chosen for reasons of efficiency, owing to the nature of the study which was conducted in the field.
Additional Measures
After completing the forgoing key measures, participants reported their gender and age. Participants put their completed surveys in a sealed envelope and placed it in a mailbox that could not be opened by the researchers. After doing so, they received a standard funnel debriefing (Bargh & Chartrand, 2000) designed to probe whether participants were aware of the experimental manipulations, hypotheses, or their own mimicry behavior. Participants revealed neither suspicion regarding the validity of the roles played by the research assistants, nor the hypotheses, and were not aware of their own mimicry behavior (or lack thereof).
Results and Discussion
Scale Reliabilities
In the present study, conducted in the Netherlands (in English), one of the (reversed) narcissism items, “I feel embarrassed if someone compliments me”, did not correlate with any other narcissism items (inter-item correlation = .086), and was therefore not included in the calculation of mean narcissism (α = .64). For the same reason, one item (“Most people who get ahead in the world lead clean moral lives”; inter-item correlation = –.18) was also excluded from the calculation of mean Machiavellianism (α= .69). The psychopathy scale items all demonstrated good inter-item reliability (α = .70). Therefore, the final narcissism scale included eight of the original nine items, the final Machiavellian scale included nine of the original ten items, and the original nine psychopathy items were retained. Consistent with previous research (Paulhus & Williams, 2002), participants' narcissism subscale scores correlated with their psychopathy scores (r(91) = .32, p < .002), and to a lesser extent with their Machiavellianism scores (r(91) = .18, p = .08).
Status Manipulation Check
Participants’ impressions of the PhD researcher and undergraduate assistant were tested using a 2 (Researcher status: ‘PhD researcher’ vs. ‘undergraduate assistant’) × 2 (participant gender: male vs. female) × 3 (evaluative judgments: intelligent, likable, attractive), mixed-model analysis of variance (ANOVA) with evaluative judgments as within-subjects measures. This analysis revealed a main effect of researcher status on judgments of the researchers’ intelligence, F(1, 89) = 9.77, p = .002; likability, F(1, 89) = 6.46, p = .013; and attractiveness, F(1, 89) = 5.69, p = .019. Participants evaluated the PhD researcher as more intelligent (M = 5.89, SD = .93) than the undergraduate assistant (M = 5.55, SD = .86), but less likable (M = 5.41, SD = .89) and less attractive (M = 4.71, SD = 1.29) than the assistant (M = 5.59, SD = .80 and M = 5.05, SD = 1.09). We found no interaction effects of participant gender on evaluations of the experimenter and assistant, Fs < 1.70. The manipulation of researcher status was, therefore, successful in the sense that it influenced perceptions of researcher intelligence as intended.
Identification of Narcissists
In order to identify participants with extreme differences in narcissism scores (M=2.92; SD=.48), we examined those whose narcissism scores were 1SD above the mean (high narcissists) and those who scores were 1SD below the mean (low narcissists). In this way, 12 narcissistic (5 male, 7 female) and 11 non-narcissistic participants (4 male, 7 female) were identified.
Narcissism, Status, and Mimicry
The hypothesis being tested in the present research is that the nonconscious mimicry behaviour of narcissists, a group of individuals defined by their desire to be liked by high status others, will increase as a function of the status of the interaction partner. A 2 (Researcher status: ‘PhD researcher’ vs. ‘undergraduate assistant’) x 2 (narcissism profile: low vs. high narcissism) repeated measures ANOVA revealed a no main effects of researcher status or narcissism profile on mimicry, Fs < 1.0.
As predicted, however, there was a significant interaction between interaction partner status and the narcissism of the participant, F(1, 21) = 7.84, p = .011, pη2 = .27. 1 As presented in Figure 1, the mimicry behaviour of below-average narcissists was not contingent upon the status of the interaction partner; participants identified as being below-average in narcissism mimicked the ‘undergraduate assistant’ (M = .36, SD = .67) and the ‘PhD researcher’ (M = .09, SD = .30) equally, F(1, 21) = 1.87, p = .19, pη2 = .08. By contrast, the mimicry behaviour of above-average narcissists was contingent upon the status of the interaction partner: Participants with above-average narcissism scores mimicked the higher-status interaction partner more (M = .67, SD = .78) than the lower-status interaction partner (M = .17, SD = .39), F(1, 21) = 6.86, p = .016, pη2 = .25.
The impact of individual narcissism (above or below average) on mimicry behavior depends on the status of the interaction partner (high vs. low). Error bars represent 1 standard error above and below the mean. The Y-axis reflects the range of mimicry behavior demonstrated by the sample (0–3).
Further analyses revealed that while the lower-status ‘undergraduate researcher’ was mimicked to the same extent by participants with below-average (M = .36, SD = .67) and above-average narcissism (M = .17, SD = .39), F(1, 21) = .75, p = .40, pη2 = .04, the higher status `PhD researcher' was mimicked significantly more by participants with above-average narcissism (M = .67, SD = .78) compared to those with below-average narcissism (M = .09, SD = .30), F(1, 21) = 5.27, p = .032, pη2 = .20.
In summary, participants with above-average narcissism scores (controlling for Machiavellianism and psychopathy) mimicked a high-status other more than a low-status other, and mimicked the high-status other more than participants with below-average narcissism (controlling for Machiavellianism and psychopathy.
Additional Analyses
Findings regarding the relationship between liking and mimicry are mixed (Chartrand & Bargh, 1999; Yabar, Johnston, Miles, & Peace, 2006). In the present research, with respect to all participants surveyed (N = 91), there was no correlation between liking e.g., and mimicry of the `PhD researcher' (r = -.009), nor liking for and mimicry of the `undergraduate assistant' (r = .07), even when controlling for attractiveness and intelligence. Among selected participants with above or below average narcissism residual scores, this pattern remained (ps > .80).
General Discussion
Past research has coined behavioral mimicry as a nonverbal signal of the intention to “tend and befriend” (e.g., Lakin, Jefferis, Cheng, & Chartrand, 2003). The present research suggests that mimicry is not always a reliable cue to a person’s social motives; that under certain conditions, certain personalities may exhibit behavioral mimicry in response to a desire to be liked in the absence of genuine liking. Specifically, we found that narcissists who have a chronic desire to be liked by higher status others, exhibited more behavioral mimicry of higher status than lower status others. Further, despite reporting less affection for the higher status interaction partner than did non-narcissists, narcissists displayed significantly more mimicry of these higher-status others than non-narcissists. Hence, the “wolves” (narcissists) wore “sheep’s clothing” (displayed behavioral mimicry) in response to a desire to be liked rather than actual liking.
While the present investigation of the influence of the desire to be liked on mimicry was rooted in the observation of a specific class of individuals, the desire to be liked is a goal that may also be situationally induced in almost anyone (except perhaps Machiavellian and psychopathic personalities). People are motivated to self-present in a variety of different situations (e.g., first dates, job interviews) and in interaction with a variety of different types of people (e.g., one’s boss, or someone particularly influential or even threatening). The present research suggests that a strong desire to be liked may increase behavioral mimicry in these situations and interactions. Future research will explore the variety of situations in which mimicry may be elicited by a self-interested desire to be liked rather than actual liking.
Beyond offering a new perspective on the types of motivations that can facilitate the expression of automatic behavioral mimicry, the within-and between-subjects design of the present research demonstrates that mimicry behavior varies both between individuals and across situations. Despite being an automatic process, therefore, behavioral mimicry is highly sensitive to individual motives and situational contingencies. In this respect, the present research contributes to the emerging perspective that mimicry is a strategic, albeit nonconscious social behavior (Lakin, Chartrand, & Arkin, 2008).
By observing that the mimicry behavior of narcissists is sensitive to social status cues (intelligence), the current research suggests that differences between the mimicry behavior of narcissists and non-narcissists may be due to situation-specific differences in the desire to be liked as opposed to individual differences in affect or cognition. Further support for the notion that the desire to be liked is sufficient to increase the expression of behavioral mimicry is needed, however, particularly in view of the use of a personality differences in “desire to be liked” rather than a direct experimental manipulation and measurement of “desire to be liked” to investigate the effects of this motivation on mimicry behavior.
The present research also adds to our understanding of narcissism. One of the most interesting paradoxes of narcissism is that it is associated greater popularity with acquaintances—the offensive nature of narcissists only becomes apparent with greater familiarity (Paulhus, 1998). Back, Schmuckle, and Egloff (2010) recently explored why narcissists are so popular at first sight, and found that narcissists look and behave in ways that create a positive impression (e.g., wearing fancier clothes, charming facial expression, more self-assured body movements, and more verbal humor). Adding to this, the present research suggests that some narcissists may also win over their audience by more frequently mimicking others.
One of the boundary conditions of the current findings is that narcissists are only charming as long as their ego is not directly threatened (Morf & Rhodewalt, 2001). In the present research, narcissistic participants discussed everyday life events (as opposed to a topic with which the participant was not familiar or did not have expertise) with a “PhD researcher” for a few minutes in the absence of direct competition for social standing. In the absence of ego-threat, narcissists mimicked higher status others, theoretically in pursuit of self-affirmation and status by association. If the narcissist interacted with a higher status other in a context that threatened their self-esteem, however, we would expect to see hostile nonverbal behavior rather than mimicry, as narcissists have been shown to display self-defensive aggression toward those who pose a threat to their fragile self-esteem (Morf & Rhodewalt, 2001).
In conclusion, mimicry may be increased by desire to be liked rather than actual liking. Specifically, the present research found that despite not liking higher status individuals any more than lower status individuals, they mimicked them more frequently. Since narcissism is characterized by a desire to be liked by higher status others in order to protect their inflated self-views, the status-contingent mimicry behavior of narcissists suggests that mimicry may be facilitated by a desire to be liked in the absence of actual liking.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.