Abstract
The drag queen cultural phenomenon has been described at length. However, the depiction of outlandish and hyperbolic womanhood and taunting and formidable behavior at the core of drag queens’ public persona has still to be fully accounted for. We argue that these aspects of the drag queen’s public appearance could best be understood in a signaling framework. Publicly donning extravagant woman’s costumes attracts harassment and brings financial, mating, and opportunity costs, generating the conditions for the transmission of honest signals. By successfully withstanding those odds, drag queen impersonators signal strategic qualities to members of the gay community. Data collected among gay and straight participants support a costly signaling reading of the drag queen cultural phenomenon. Participants generally agree that successful drag queens typically incur costs, while gaining specific social benefits.
The drag queen phenomenon has drawn much attention over the past decades (Berkowitz & Belgrave, 2010; Berkowitz, Belgrave, & Halberstein, 2007; Hopkins, 2004; Newton, 1972; Rupp, Taylor, & Shapiro, 2010; Sagarin, 1969; Schacht, 2000, 2002; Schacht & Underwood, 2004; Taylor & Rupp, 2004; Tewksbury, 1993, 1994). Much of the research has focused on identity-based and sociopolitical motivations to perform in drag, such as critique of traditional gender roles (Brubach & O’Brien, 1999; Garber, 1997; Rupp & Taylor, 2003), and queer political militancy (Feinberg, 1996; Muñoz, 1999; Schacht, 2000; Taylor, Rupp, & Gamson, 2005). We propose an additional, and potentially alternative, signaling account of the phenomenon’s characteristic traits of hyperbolic depiction of womanhood and formidable and exaggerated behavioral expressions (Berkowitz et al., 2007; Friedman & Jones, 2011; Schacht, 2002; Taylor & Rupp, 2004).
In a landmark publication on the life of drag queens, Newton (1972) emphasized the conspicuous, confrontational, territorial, and effeminate behavior of drag queens and the discrimination, harassment, and stigmatization that impersonators regularly had to face from both gay community members and outsiders. Most people, including the impersonators themselves, seemed to view the drag queen attitude as extreme and particular (Newton, 1972, p. 6). Despite this stigmatization, participation in the drag subculture appeared to have afforded jobless, young, and poor gays some opportunity to distinguish themselves from lower status individuals such as hustlers or “freaks,” and, for the most successful drag queens, a chance to develop celebrity-like status and social might in the gay community (Newton, 1972, p. 6).
Signaling theory has provided a theoretical framework for better explaining evolutionarily puzzling human behaviors (Bird, Smith, & Bird, 2001; Sosis & Bressler, 2003). We propose to analyze the phenomenon of drag queen behavior in light of signaling theory. Despite the costs involved in publicly endorsing a drag queen persona, marginalized individuals might find it attractive, given the benefits they stand to gain such as an enhanced reputation and increased social capital (e.g., Newton, 1972; Hopkins, 2004). The drag queen phenomenon provides an interesting case study where particular behavioral signals enhance individuals’ reputation and welfare, while being entirely decoupled from any reproductive payoff. The phenomenon can be understood as the partial output of universal cognitive mechanisms for status seeking and partner seeking. Typical organizational features of the gay community also play a role in the emergence of the drag queen cultural practice.
Background
The Drag Queen Social World
Drag queens, or female impersonators, differ from transsexuals and individuals with transvestic fetishisms 1 in that they are gay individuals who don female clothing with the explicit goal of performing in front of audiences (Schacht, 2000). The focus of our discussion is on the behavior of male impersonating female 2 rather than its opposite, female impersonating male, known as drag kings, which are part of a recent phenomenon with specific features and rationale that differ from its male counterpart (Rupp et al., 2010). Cross-gender impersonators in the lesbian, gay, bisexual, transgender, queer (LGBTQ) community are predominantly male (Rupp et al., 2010). Drag queens explicitly distance themselves from individuals who use either hormones or undergo gender reassignment surgery. These individuals are considered as not “playing fair” and are banned from participating in drag queen competitions (Berkowitz et al., 2007; Hopkins, 2004; Schacht, 2002; Taylor & Rupp, 2004).
Individuals who endorse a drag queen public persona are generally in their late teens to 40s, with an average age in the late 20s (Berkowitz & Belgrave, 2010; Berkowitz et al., 2007; Tewksbury, 1993). The available research, based on small sample sizes, indicates that most female impersonators have achieved a low educational level, and their earnings are at the lower end of the income distribution (Berkowitz & Belgrave, 2010; Newton, 1972). Female impersonators’ subdued demeanor when not in costume has often been noted in ethnographies of the drag queen scene (Friedman & Jones, 2011; Strübel-Scheiner, 2011; Tewksbury, 1993). They have also been described as “not of the prescribed standard of gay beauty” (Strübel-Scheiner, 2011, p. 12). The main reasons mentioned in the literature for becoming a drag queen include the desire of raising one’s social standing, getting more involved in the gay community, and starting a career in entertainment (Hopkins, 2004; Taylor & Rupp, 2004; Tewksbury, 1993, 1994).
Drag queens don their costumes primarily to perform at gay bars, nightclubs, and organized competitions (Berkowitz et al., 2007; Schacht, 2002). Their apparel is not intended to depict ordinary female attires like that of transgender women but portray purposefully outlandish, often vulgar, and exaggerated stereotypes of womanhood (Harris, 1995; Tewksbury, 1994). Theatrics, choreographed dance, and lip-syncing to short popular musical numbers are part of club and bar shows (Berkowitz et al., 2007; Schacht, 2002). Performers earn some income thanks to spectators’ tips. Only the most successful drag queens can generate enough money to fully support their livelihood (Berkowitz & Belgrave, 2010; Berkowitz et al., 2007; Hopkins, 2004; Newton, 1972; Schacht, 2002). Female impersonators progressively build reputation and, for some, local fame, through patronizing and performing at bars and nightclubs (Hopkins, 2004; Tewksbury, 1994). Outside of cabaret performances, the public wearing of drag queen attire is usually restricted to gay festivals or exceptional events (Taylor & Rupp, 2004). Of note, however, at the beginning of the fight for gay rights in the 1960s, many less successful drag queens (aka street drags) would parade ostentatiously adorned through the streets, often attracting the ire of passersby (Newton, 1972).
The literature depicts a typical portrait of a successful drag queen’s public persona made of playful rudeness (Friedman & Jones, 2011), obnoxiousness (Taylor & Rupp, 2004), and unkind critical spirit (Friedman & Jones, 2011). Drag queen role-play typically involves aggressive teasing and mocking of audiences (Berkowitz et al., 2007; Schacht, 2002; Taylor & Rupp, 2004). Boisterous displays comprising “flesh flashing” and focus on the male–female duality of the drag queen persona are core components of public performances (Harris, 1995; Rupp et al., 2010; Schacht, 2002; Taylor & Rupp, 2004). Female impersonators aggressively joke about their masculinity, at times calling attention to their tucked away penis, emphatically affirming that they are “chicks with dicks” (Berkowitz et al., 2007; Rupp et al., 2010; Taylor & Rupp, 2004). Their sustained involvement in public confrontations has led to the characterizations of drag queens as militants, warriors, and urban street fighters (Duberman, 2013; Feinberg, 1996; Harris, 1995; Schacht & Underwood, 2004).
Individuals wishing to become recognized female impersonators face many burdens. The financial expenses are significant. Costumes and makeup worth hundreds of dollars require sustained investments, seldom offset by the earnings from competitions or audience members’ tips (Hopkins, 2004). Opportunity costs are also high: successful performance depends on extensive rehearsal and sustained investments in self-promotion (Berkowitz & Belgrave, 2010; Hopkins, 2004; Schacht, 2002). Ethnographic accounts support the idea that these investments compete with maintaining more personal relationships, which probably explains why female impersonators appear to face difficulty establishing enduring romantic relationships (Berkowitz et al., 2007). Furthermore, costumes and makeup constitute potent sexual turnoffs, cutting off drag queens from the available mate pool when engaged in public display of their persona (Harris, 1995). They also face increased risk of being harassed and discriminated against given the heightened social visibility to which their trade exposes them (Berkowitz & Belgrave, 2010; Berkowitz et al., 2007; Friedman & Jones, 2011; Hopkins, 2004; Newton, 1972).
Among the people who attempt to join the drag queen community, few become great successes. Amateurs have limited status recognition and are exposed to stinging criticism, mockery, and gossip of their gay audiences (Berkowitz & Belgrave, 2010; Schacht, 2002). Beginners face wholesale ridicule for their unrefined appearance and behavior (Hopkins, 2004). Scrutiny from other drag queens is intense and competition for recognition between impersonators is great (Berkowitz & Belgrave, 2010; Friedman & Jones, 2011; Hopkins, 2004). Few paid professional opportunities are available at gay establishments, intensifying rivalry (Hopkins, 2004). In some communities, the intense competition facilitates the emergence of “drag families,” with seniors controlling and mentoring cadets (Hopkins, 2004). These families do help in reducing the costs of entry for beginners. However, impersonators who have support of such families generally do not garner as much respect as those who succeed independently (Hopkins, 2004).
Contrasted with the intense pressure placed on amateurs, successful impersonators end up commanding some authority in the gay community (Berkowitz & Belgrave, 2010; Berkowitz et al., 2007; Hopkins, 2004). Since many drag queens begin with a low social endowment and status, gains in social capital in the gay community can be significant. They can eventually earn greater respect, becoming role models for up-and-coming queens, and might be sought after to animate social events and parties (Berkowitz et al., 2007; Hopkins, 2004). Such successful drag queens can make enough money from club bookings to support a full-time career (Berkowitz & Belgrave, 2010). These social and economic incentives make becoming a drag queen an attractive venture for at-first-marginalized low-status individuals in the gay community, who may not have easier alternate opportunities to gain social acceptance and recognition.
Emergence of the Modern-Day Social Phenomenon
Early descriptions in existing literary sources placed the first manifestations of drag queens, akin to their modern-day instantiations, in England of the early 1700s (Norton, 2016). At the time, gay sexual activities were illegal, systematically prosecuted, and could result in capital punishment. Performances were strictly restricted to audiences of gay men in meeting places called “molly houses” (Norton, 2016). Most of what is known about these drag queens was discovered from raids on molly houses and court proceedings (Norton, 2016).
The earliest appearances of gay female impersonators in America recorded in the literature happened at masquerade balls held in New York City in the early 20th century (Chauncey, 1994). Such masquerades allowed individuals to connect and to engage in same-sex courtship without being directly exposed to social sanctions. In the early 1930s, there were several recorded instances of “public disturbances” involving female impersonators parading through the streets of New York (Chauncey, 1994).
Out of the underground gay social networks of the early 20th century, predominately gay neighborhoods progressively emerged. Gay-only communities were well on their way to consolidation by the 1960s (Harry & DeVall, 1978). Gay bars in those neighborhoods became prime locations to find same-sex romantic and sexual partners and to build social support networks (Chauncey, 1994; Harry, 1974; Hooker, 1965). These neighborhoods provided a certain level of safety for their residents during the McCarthy era when gay people were purged from government jobs, and gay establishments were subjected to intense police surveillance and raiding (Loftin, 2007).
To avoid detection and the risks of harassment, job loss, or arrest, gays publicly conformed to straight mannerisms. “Swishes,” feminine acting gays drawing attention from law enforcement, were openly despised (Loftin, 2007). The risk was aggravated for gender nonconformists such as female impersonators (Arriola, 1995; Chauncey, 1994; Loftin, 2007). It was during this postwar period that drag queens emerged as significant players in the gay community (Arriola, 1995; Hillman, 2011). In the 1960s, they became very visible during the violence opposing police and gay communities in both San Francisco and New York City (Arriola, 1995; Hillman, 2011). It is noteworthy that many of the female impersonators who became prominent in the 1950s and 1960s were from very poor backgrounds (Hillman, 2011; Newton, 1972).
Cross-culturally, drag queens are now found in gay communities throughout the Western world. Interestingly, the phenomenon has also appeared in regions where gay individuals face important social or legal persecution such as in Moscow (Kirk, 2015), Jerusalem (Middle East Eye, 2016), and parts of Africa (The Independent, 1999). Other cross-cultural examples such as the Khwaja Sira of Pakistan (Khan, 2014) or the Fa’afafine of Samoa (Vasey & Bartlett, 2007) share some similarities to the drag queen phenomenon (e.g., harassment, ostracism, and hyperbolic displays of femininity). However, those cultural manifestations have their respective specificities that differentiate them from the modern-day Western drag queen phenomenon. We focus solely on the latter.
Signaling Theory as an Explanatory Framework for the Drag Queen Phenomenon
Signaling theory has explained a wide variety of evolutionarily puzzling and, in appearance, extravagant behavior in animal species (Zahavi & Zahavi, 1999). The theory holds that in specific conditions, an extraordinary (i.e., apparently nonfunctional) behavior that communicates information about the sender can become evolutionarily stable if, on average, it conferred fitness benefits to its sender (Maynard-Smith & Harper, 2003; Zahavi & Zahavi, 1999). According to the handicap principal, such evolutionarily stable signal must have a specific characteristic to be reliable: it should be costly, rendering it hard to fake, a requirement if it is to be perceived as indubitable signal carrying honest information about its sender (Zahavi, 1975, 1977).
All signals, as opposed to signs, are manipulative, being designed to modify receivers’ behavior (Dawkins & Krebs, 1978; Maynard-Smith & Harper, 2003). Signals therefore have an intended audience (e.g., potential mates, cooperative allies, and competitive adversaries). Whether or not a signal is effective in manipulating the behavior of others depends on the signal being detected by its intended receivers and the latter’s skepticism toward the honesty of the signal (Soler, Batiste, & Cronk, 2014).
Humans use a host of costly signals to communicate honest information about themselves. For example, Meriam males signal their leadership and organizational abilities by leading boat crews in quest for food sources in high variance in the environment, such as sea turtles, instead of foraging, which would be less risky while providing much greater caloric return (Bird et al., 2001). By organizing such turtle hunts, Meriam hunt leaders convey information about their abilities, skills, and potentials to prospective mates or cooperative partners. Benefits conferred to the signaler, such as increased mating opportunity or social and political influence, may pay off in increased overall fitness (Bird et al., 2001). Other well-described examples include, Ache men’s signaling of cooperativeness by sharing more food than they receive from others (Gurven, Allen-Arave, Hill, & Hurtado, 2000), the purchasing of luxury brands as a signal of wealth and status (Nelissen & Meijers, 2011), and costly prorelationship acts as signals of commitment to romantic partners (Yamaguchi, Smith, & Ohtsubo, 2015).
Apparently irrational, extravagant, or risky behaviors, at times, also might serve a signaling function when, for instance, information about the quality of an agent is hard or costly to acquire or to infer in noisy environments. In these circumstances, it might be beneficial for individuals to strategically emit indubitable signals of their qualities to potential partners. A risky behavior, given its inherent costs, might provide for an adequate honest signal. As Fessler, Tiokhin, Holbrook, Gervais, and Snyder (2014) have shown, deliberate risk-taking increases perceptions of formidability, with men engaging in such behavior typically perceived as larger, stronger, and more violent than others. However, formidability is likely to be signaled by more than just sheer strength and observable physical attributes. Other factors such as cues of resoluteness and size of social support also increase an agent’s perceived formidability (Fessler & Holbrook, 2013; Holbrook & Fessler, 2013).
Signaling one’s formidability in particular social environments, such as those characterized by conflict and competition, is likely to be beneficial for deterring potential adversaries and for attracting potential allies. Gambetta (2009) found that prisoners often use deliberate self-harm to signal fierceness and fearlessness in the highly constrained social environment of correctional facilities. Displays of self-harm are particularly common in prison environments with high inmate turnover and unstable social hierarchies. Such signaling allows individuals who are not naturally physically formidable to broadcast to large audiences a disposition for fierceness approximating extreme irrational impulsivity, hence affording them some protection from aggression in the closed prison world. Martinez and Lienard (2015) in a survey of study cases in the Human Relations Area Files (eHRAF World Cultures, 2014) found that institutionalized public displays of deliberate self-harm are systematically associated with social environments characterized by low social mobility and competition between marked and ranked social units. Participants are more likely to be young adult males who lack access to other more ordinary status enhancers, such as wealth, and human and social capitals.
The gay community has historically shared many of the characteristics of social worlds where costly signaling can emerge. It is a community that has been ostracized from mainstream society and has responded by a self-imposed isolation to escape persecutions. In some respects, the gay community constitutes a protected social world with its own organization, rules of courtship and socialization, behavioral norms, and an exacerbated mating and status competition (Bailey, Gaulin, Agyei, & Gladue, 1994; Chauncey, 1994; Kenrick, Keefe, Bryan, Barr, & Brown, 1995). 3 Youth, physical attractiveness, and potential for uncommitted sex play an essential role in mate searching (Bailey et al., 1994). An increased interest in younger partners correlates with seekers’ age (Kenrick et al., 1995). These mate preferences probably contribute to the observed enhanced competition in the gay mating market. Of note, the gay community has a high prevalence of body dissatisfaction and eating disorders, supporting the view of an intensely competitive mating landscape (Kaminski, Chapman, Haynes, & Own, 2005; Russell & Keel, 2002; Siever, 1994; Tiggemann, Martins, & Kirkbride, 2007). On average, gay men also prefer men who describe themselves as masculine rather than feminine (Bailey, Kim, Hills, & Linsenmeier, 1997), and feminine gay men regularly report discrimination from other gay men (Taywaditep, 2001). The drag queen phenomenon gains to be understood against such general backdrop of mate preferences and highly competitive mating market. The drag queen persona combines both the donning of particular attires (i.e., strictly speaking, the handicap, which is, the exaggerated feminine appearance) and a typical behavior (i.e., the stance, assertive, and provocative). The antithesis, namely, the association of aggressive provocation and exaggerated feminine markers, constitutes the entirety of the signaling.
As previously mentioned, beginner drag queens are younger individuals often lacking more traditional status-enhancing means such as attractiveness, education, and wealth (Berkowitz & Belgrave, 2010; Newton, 1972; Strübel-Scheiner, 2011). There is significant evidence that by donning their costumes and makeup drag queens increase their risks of being exposed to harassment and humiliation, violence or ostracism, loss of mating opportunities, and associated opportunity costs (Berkowitz & Belgrave, 2010; Berkowitz et al., 2007; Hopkins, 2004; Newton, 1972; Schacht, 2002). However, by engaging in daring nonconformist practices, impersonators might at the very same time achieve some recognition, as successful performances can be perceived as public displays of one’s claim to possess the ability to take on any associated challenge. Impersonators communicate resoluteness and fearlessness, in brief, a type of social formidability. In competitive environments, formidable individuals are able to bargain for their own welfare and attract coalitional allies more easily than less formidable agents (Sell, Tooby, & Cosmides, 2009). For low-status entrepreneurial individuals, with little to lose, unwilling to accept the ranking ascribed by some of their observable attributes, the potential payoffs of risk-taking behaviors should be very attractive.
Thus, the hyperbolic and extravagant expressions of womanhood characteristic of drag queen personas call attention to extraordinary behaviors and the associated costs that come with it. Its typical confrontational attribute is a prominent reminder of some type of formidability. The historical record of drag queens’ participation as lead figures during prominent 1960s countercultural riots supports the characteristic of pugnacity associated with a drag queen persona (Arriola, 1995; Hillman, 2011). Portraits of drag queens as militant risk-takers are also readily documented in contemporary accounts (Nichols, 2013; Paul, 2014).
In order to assess whether signaling theory is an appropriate model for understanding the drag queen phenomenon, we need to address four main questions: (1) Do people perceive the typical features presented in the literature on the drag queen persona? (2) Is the signaling dimension of the impersonation better perceived by gays? (3) Are gays overall more knowledgeable about the drag queen cultural practice? and (4) Are the costs involved in impersonating understood? These questions have not been explored in the literature yet. We relied on a survey measure distributed to members of the gay community and of the general public to explore these matters.
Method
Participants
We used two participant pools: (1) a sample of members of the gay community recruited through online social media groups and (2) a general population sample reached through Amazon’s Mechanical Turk® (MTurk). The University of Nevada Las Vegas’ institutional review board approved the research. Participants were screened and excluded for incomplete survey responses (n = 55). Not being the target of this study, individuals who perform as drag queens (n = 11) were excluded from the analysis. In order to maximize the differences between the two samples of participants and to reduce the risk of potential noise, bisexuals (n = 24) were also excluded from our two samples.
Our sample from the gay community comprised 133 men, 98% self-reported that they were only attracted to men (ages 18–74, M = 29.41, standard deviation [SD] = 9.63). Individuals were invited to participate through messages posted on gay social group Facebook® pages such as I’m Gay and I love it, Gay Facebook, and Gay California. These social groups were chosen as recruitment sources because many of their members are currently active within the gay community attending gay bars and nightclubs on a regular basis. Furthermore, these online social groups allowed us to sample gay individuals from a broad range of gay communities throughout the United States. Membership of those social groups ranged from 1,200 to 92,000. Participants were entered into a drawing to win a US$50 gift card.
In order to compare the perceptions of gay individuals to those of the general population, a second sample was recruited from the Internet using MTurk website. The sample consisted of 56 (50%) males and 56 (50%) females. MTurk is a website where one posts tasks to be completed for a certain compensation. Participants recruited through MTurk are fairly representative of the U.S. population (Buhrmester, Kwang, & Gosling, 2011). Recruitment was limited to the United States. Eight participants who self-identified as gay and 16 that self-identified as bisexual were removed from the sample, leaving 112 participants (ages 18–72, M = 34.26, SD = 12.31). Of these participants, 50% reported being only attracted to females and 49% reported being only attracted to males.
These participants were compensated US$0.40 for their time. Individuals of the first and second samples had comparable levels of education, with 71% and 62%, respectively, having completed at least some postsecondary education. The respective mean ages of 29 and 34 years indicate that the majority of our participants were in prime years of mate seeking and sexual (and, for some, reproductive) activities.
Materials: Perception of Drag Queens Survey
A great deal of the drag queen research has concentrated on drag queens and their perceptions of the gay community (e.g., Taylor & Rupp, 2004) or on the interaction between gays and drag queens (e.g., Berkowitz et al., 2007). One exception to this is a recent study on hypermasculinity and drag queen stereotypes (Bishop, Kiss, Morrison, Rushe, & Specht, 2014). We expanded on the objectives of this research in investigating to a greater extent gays’ perception of drag queens. It should be noted that we are not interested in the personal experience of drag queens; rather, we focus on the communicative aspect of their behavior. We are also not attempting to test a specific model; instead we use an exploratory approach to show how our framework can help make sense of aspects of the phenomenon previously described in the literature. For the purpose of our research, we specifically designed a set of original survey instruments. Participants evaluated 47 statements (see Table A1) focused on five categories (Table 1).
Foci of the 50 Questions.
Procedure
Participants were directed to the Qualtrics survey website after clicking on the hyperlink at the bottom of the online advertisement. They gave their consent to participate by checking a radio button before accessing the survey. Participants were first asked three free response questions: “Have you ever thought of performing in drag? If so, what are some of the reasons for why you considered performing?” “Why do you think most drag queens perform?” “What is your definition of a drag queen?”
In order to ensure that participants properly understood the concept of drag queen, the following definition from the Oxford dictionary (Drag queen, 2014) was provided immediately after they had answered the three previous questions: A homosexual man who dresses as a woman especially to entertain people.
Participants were then instructed to read the following sentences: Take a moment to think about drag queens in your community. What, do you think, are some of their important traits? Please indicate how much you agree or disagree with the following statements as they pertain to drag queens. Participants evaluated a series of 43 statements plotting their response on a 7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree). Statements included, for example, drag queens are more trustworthy than others in the gay community, and drag queens must be more blunt than others in the gay community. By asking participants to respond in this manner, we were looking for stereotypic responses. Stereotypes are cognitive structures that can be useful in understanding how individuals calibrate and adapt their behavior when interacting with others (Fiske & Neuberg, 1990). Some questions such as is there such a thing as a fake drag queen? allowed for free response once the participant responded in the affirmative. In the present case, when the participant’s answer was positive, he was directed to a new screen where a short answer question box appeared, asking the subject, what would make a fake drag queen? Given that the participants from our MTurk sample (i.e., general population) might not have been familiar with the drag queen phenomenon, participants were instructed to answer the questions to the best of their knowledge. Following the survey questions, participants responded to more general questions about their involvement in the gay community, such as how often they attend gay nightclubs or bars, and additional general demographic information.
Results
Knowledge of the General Phenomenon
Given the heterogeneous covariance matrices between gay and straight samples when examining the full set of variables, a quadratic discriminant analysis was performed to determine whether the items derived from the literature discriminated between the responses of the gay and straight participants. The reclassification of cases was highly successful: 98.33% of the original grouped cases were correctly reclassified into their original categories and 74.17% of cross-validated cases were correctly classified (Table 2). These results indicate that the gay sample has a response pattern that distinguishes them from the straight sample participants. In addition, Mann–Whitney U tests indicated that there were significant differences between our straight and gay sample for 28 of the 43 variables (65%; see Table 3).
Quadratic Discriminant Analysis Classification Outcomes.
Note. 98.33% of original grouped cases are correctly classified. 74.17% of cross-validated cases are correctly classified. Five participants are removed listwise from analysis. Prior probabilities are proportional to group size.
Survey Item Means and Mann–Whitney U Test Results Ordered by Significance.
Note. SD = standard deviation.
*Mann–Whitney U significant at the p < .05. **Mann–Whitney U Significant at the p < .01.
Stereotypic Features of the Phenomenon
We wanted to examine whether participants in both our samples perceive the typical attributes of the drag queen persona that is described in the literature. Items with the strongest level of agreement (>70% agreement) after combining responses (strongly agree to somewhat agree/strongly disagree to somewhat disagree) within the gay and the straight samples were selected for further analysis leaving 18 variables (Table 4).
Items With High Levels of Agreement.
Note. Any value ≥70% appears in boldface. A% = agree; D% = disagree.
Discriminant analysis determined which of these variables best predicted membership in the straight or gay sample. Previous research found discriminant analysis to be robust when analyzing skewed data (Sever, Lajovic, & Rajer, 2005). The overall χ2 test was significant, Wilks λ = .734, χ2(18) = 70.13, R 2 = .516, p < . 001. Reclassification of cases based on the new canonical variables was successful: 70% of the cases were correctly reclassified into their original categories. Table 5 shows the independent variables, which are strongly associated with discriminant function 1 (r > .3) and distinguish between participants from our straight sample and gay sample. Items that did not discriminate well between the two groups included: Individuals who perform in drag face potential physical assault from straight people (r = .168), Individuals who perform in drag face potential ridicule or insult from straight people (r = .136), and Individuals while performing in drag sexually arouse me (r = −.048).
Structure Matrix.
Participants in the gay sample appear to have greater knowledge of certain aspects of the drag queen phenomenon. We wanted to further explore within the gay sample what features are most clearly recognized. To determine the number and nature of factors underlying the stereotypic features of the drag queen impersonator in the gay sample, a factor analysis was conducted on the 18 remaining items from Table 4. Kaiser–Meyer–Olkin measure of sampling adequacy was .69, above the recommended value of .6, and Bartlett’s test of sphericity was significant, χ2 (120) = 1,000.12, p < .001, suggesting that the data were suitable for analysis (Bartlett, 1954; Kaiser, 1974). Parallel analysis (Horn, 1965; Cota, Longman, Holden, & Rekken, 1993) and analysis of the scree plot suggested that a four-factor solution was optimal (see Figure B1 for scree plot). Polychoric, as opposed to Pearson, correlations were used as the univariate skewness and kurtosis values suggested that the variables were not normally distributed (Holgado–Tello, Chacón–Moscoso, Barbero–García, & Vila–Abad, 2010). The method of factor extraction used was a principal axis factoring with oblimin rotation. Variables that had a factor loading less than .3 were removed from analysis, leaving 16 variables. Using the polychoric correlation matrix, the ordinal α for each of the four factors was calculated as .85, .91, .81, and .65, respectively (Gadermann, Guhn, & Zumbo, 2012). Overall, the four factors accounted for 53.8% of the variance. It should be noted that the purpose of our factor analysis is exploratory. Factors 2 and 3 contain only two loadings because of the limited number of items asking about mating and aggression costs, and the α for factor 4 is less than .7, hence the results of our factor analysis should be interpreted with caution.
Composite scores were created for each of the four factors, by computing grand means on the base of the means of all items loading onto their respective factor. Review of the factor loadings (Table 6) suggested that Component 1 showed a strong positive association with social eminence and influence and was labeled social formidability (M = 5.41, SD = 0.89). Component 2 had a positive association with sexual attraction to drag queens and was labeled mating cost (M = 1.79, SD = 1.14). Component 3 was related to potential social and physical harm that drag queens face and was labeled risk of aggression (M = 5.87, SD = 0.99). Component 4, showing an association with social investment and capital, was labeled social investment (M = 5.61, SD = 0.72).
Principal Axis Factor Analysis.
Note. h 2 = communality. Rotation method: direct oblimin. Salient factor pattern matrix coefficients >.35 are bolded. Factor 1 = social formidability; Factor 2 = mating cost; Factor 3 = risk of aggression; Factor 4 = social investment. Factor 2 has been reverse scored.
Discussion
Do People Perceive the Typical Features of the Drag Queen Persona? Do Gays and Straights Understand the Costs Involved in Female Impersonation?
Our findings support a communicative intent of drag queen behavior. It appears that there are universally perceived aspects of the signal, specifically related to the cost of the behavior. The gay and straight population concur that drag queens are generally less attractive as potential mates and exposed to heightened risk of aggression. Thus, both the gay sample and general population perceive some costs associated with becoming a drag queen.
We do not have data on the rate at which drag queen prospects fail. The drag queens on the base of whom people elaborate their stereotypical understanding of the phenomenon are individuals still engaged in competing and addressing the challenges they took on, that is, having some amount of success in addressing the challenges their behaviors and attire raise. The severity of the costs, which might involve such things as disrespect, hazing, harassment, and loss of mate potential, will vary according to the rank that the individual occupies from novice to established and famous drag queen. One thing seems to be essential though, for both our samples, having to withstand affronts and offenses is a core feature of the drag queen phenomenon.
Gay nightclubs and bars are places for effectively broadcasting traits and dispositions to large audience. However, the cost involved in endorsing a drag queen persona seems to be perceived widely, even by members of our nongay sample much less likely to be directly exposed to drag queens (among our straight sample 77% indicated they had never been to a gay establishment). The signal is likely to be received even more strongly by the gay community, given what gay individuals know about the cost of expressing publicly one’s nonconformist sexual orientation. In the late 1980s, up to 92% of gays and lesbians were targets of verbal abuse or threats during their lifetime because of their sexual orientation and 24% were victims of physical assault (Herek, 1989). A more recent sample found that nearly 50% had been verbally assaulted or abused (Herek, 2009). To sum up, our data show that the cost of being a female impersonator is the main factor about which most participants agree that it is a core part of adopting a drag queen persona.
Is the Signaling Dimension of the Impersonation Better Perceived by Gays?
While the general cost associated with being a drag queen is perceived by both samples, the results of the quadratic discriminant analysis indicate that the response patterns of the two samples are dissimilar, suggesting that our two pools of participants have distinct knowledge of the phenomenon. Review of the items with the highest level of agreement within the two populations shows that only 9 items reached 70% agreement within our straight sample, whereas 18 items exceeded 70% agreement within our gay sample indicating that our gay participants are much more likely to have a definite opinion than our straight participants, suggesting that certain aspects of the signal may be intended more narrowly for a gay audience.
Discriminant analysis of the items with the highest level of agreement within the two populations shows that items associated with sociality and formidably (e.g., it is not wise to upset a drag queen, successful drag queens are considered to be local celebrities in the gay community, drag queens typically have a greater social network than others in the gay community) were the best items for discriminating between the responses of our straight and gay participants. This lends support to our claim that the gay community might be the prime target of the drag queen signaling.
Are Gays Overall More Knowledgeable About the Drag Queen Cultural Practice?
The underlying structure of the stereotypic features identified by the factor analysis for our gay sample is consistent with a costly signaling model. As previously described, the cost of the behavior is clearly recognized among our gay participants. Those participants typically insist that drag queens do not constitute interesting potential mates hence, readily acknowledge the mating costs. Undoubtedly, by donning their special apparel and behaving outlandishly, female impersonators remove themselves from the mate market precisely where finding partners typically takes place, gay nightclubs and bars. Furthermore, participants understand well that drag queens increase their exposure to harassment and violence. The withstanding of those costs allows for the honest actualization of the sender’s formidability.
Drag queens are also perceived to be socially formidable and invested individuals with influence in the gay community. Being a recognized and successful drag queen means being willing to engage in aggressive behavior and endorse the role of provocateur. Among our gay participants 73% agree that it is not wise to upset a drag queen (only 5% disagreed). This statement is also the most discriminant item between our gay and straight participants. The involvement of drag queens in many early confrontations between the gay community and police forces, as well as contemporary accounts of female impersonators’ attitude at gay pride events, provides evidence of the role expectation of provocation that comes with acting the part of a drag queen (Arriola, 1995; Hillman, 2011; Nichols, 2013; Paul, 2014).
Note that the social might associated with formidability is often imposed on others. Social agents would rather have a formidable individual on their side than against them. Interestingly, the results of a recent study suggest that gays who score higher on a masculinity measure were more likely to perceive negatively valence attributes, such as melodramatic and emotional affectations, as more characteristic of drag queens (Bishop et al., 2014). The scale used in the study evaluating masculine traits included statements such as I value power over other people and I don’t mind physical violence to defend what I have; the types of dispositions that strongly correlate with formidability. Bishop, Kiss, Morrison, Rushe, and Specht (2014) suggest that males who endorse a strong “masculine” ideal may hold negative views of drag queens because they violate gender norms that they hold dear. However, it is plausible that gay individuals scoring high in masculinity perceive female impersonators as rivals competing through different means for status and social recognition.
Conclusion
The potential benefit of social upward mobility gained from displaying specific forms of formidability seems to be a major feature that makes the drag queen institution attractive to some. The social ecology of the gay community resembles other environments propitious to institutionalized displays of formidability (Gambetta, 2009; Martinez & Lienard, 2015). In that respect, it is noteworthy that the timing of the radicalization of the phenomenon coincides with an era when the gay community was once again under threat of violence and ostracism from mainstream society.
A universal disposition for status seeking likely drives the behavior of drag queens. Costly signaling allows for the honest communication of information about the qualities of the signaler, which eventually can lead to fitness gains. The sexual behavior of gay individuals is typically decoupled from reproductive fitness consequences. In this case, universal cognitive mechanisms motivate status- and partner-seeking behavior regardless of any potential reproductive fitness payoff.
The competition between drag queens signaling their formidability facilitates the emergence of a hierarchy. Drag families are the outcome of this self-organization of groups of competitors. Drag mothers, or successful drag queens, at the top of the hierarchy, benefit from their elevated social status. Social influence, authority, status, and increased social capital are gained the higher one climbs the ladder of success. The most notable drag queens may also eventually gain greater access to sexual partners, however, we currently lack the data to ascertain that possibility.
Our research adds to the evidence that supports a costly signaling account of the emergence of the drag queen phenomenon. However, some limitations of our research will have to be addressed in the future. We relied on online sampling for collecting our data. While our samples appear relatively diverse, they may not be representative enough of the gay or straight population. Furthermore, our research focus was on signal receivers rather than on the signalers themselves. Data on drag queen demographics, benefits obtained, and actual rate of violence to which they are exposed (compared to other members of the gay community) need to be collected to strengthen a costly signaling account of the phenomenon. Further, experimental study of the motivations for engaging in drag queen behavior is also required.
Given the intriguing historical legacy of the drag queen phenomenon, we may wonder what will be its future trajectory? Our findings leave us with additional questions. As acceptance of LGBTQ people from mainstream society increases, thereby relaxing the costs of advertising one’s sexual orientation in public, will a drag queen persona no longer be costly enough to allow social outliers to gain capital by adopting it? As costs continue to be reduced over time, will the drag queen phenomenon be reinterpreted into simply a theatrical vestige of the gay community’s former conditions? The phenomenon may indeed become more folkloric than functional. A similar trajectory was followed by the pride parades: Although gay pride parades were once serious platforms for demands of political and social changes, they are now large inclusive festive parties.
Footnotes
Appendix A
All Survey Items.
| Items on 1–7 Likert-type response scalea | |||||||||||
| 1 | It is wise to not upset a drag queen | ||||||||||
| 2 | Successful drag queens are considered to be local celebrities in the gay community | ||||||||||
| 3 | Typically, drag queens have a lot of social influence in the gay community | ||||||||||
| 4 | Drag queens typically have a greater social network than others in the gay community | ||||||||||
| 5 | Queens are better connected than most individuals in the gay community | ||||||||||
| 6 | Drag queens should carefully monitor their reputation | ||||||||||
| 7 | Despite trying, some people don’t succeed in becoming drag queens | ||||||||||
| 8 | There is a great deal of competition between drag queens within the drag community | ||||||||||
| 9 | Being socially visible is critical for a drag queen | ||||||||||
| 10 | Drag queens (not dressed in drag) are likely to be more physically attractive than the average individual in their gay community | ||||||||||
| 11 | I would date someone who regularly performs in drag | ||||||||||
| 12 | For a drag queen to be successful, they must spend most of their time out at gay nightclubs or bars in drag | ||||||||||
| 13 | It takes a lot of effort, time, dedication, and willpower to become a successful drag queen | ||||||||||
| 14 | I would have sex with a drag queen while they were out of drag | ||||||||||
| 15 | Drag queens are admired in the gay community | ||||||||||
| 16 | Drag queens are respected above all else in the gay community | ||||||||||
| 17 | The influence of a drag queen in the gay community is related to the duration of their career as a drag queen | ||||||||||
| 18 | Drag queens have more access to drugs than others in the gay community | ||||||||||
| 19 | Drag queens can easily obtain sexual partners when not in drag | ||||||||||
| 20 | Drag queens are power hungry | ||||||||||
| 21 | Drag queens are territorial | ||||||||||
| 22 | Individuals who perform in drag face potential physical assault from others in the gay community | ||||||||||
| 23 | Drag queens deserve much respect | ||||||||||
| 24 | I would have sex with a drag queen while they were in drag | ||||||||||
| 25 | Drag queens are held to a higher moral standard than others in the gay community | ||||||||||
| 26 | Those who perform in drag often have long-term relationships | ||||||||||
| 27 | One of the great incentives in becoming a drag queen is to improve their social standing | ||||||||||
| 28 | Drag queens are essential in fighting for gay rights and equality | ||||||||||
| 29 | Drag queens are more trustworthy than others in the gay community | ||||||||||
| 30 | Drag queens must be more honest than others in the gay community | ||||||||||
| 31 | Drag queens put the gay community before themselves | ||||||||||
| 32 | Drag queens must be more blunt than others in the gay community | ||||||||||
| 33 | Individuals who perform in drag face potential physical assault from straight people | ||||||||||
| 34 | Drag queens are go-to-people when conflicts arise in the gay community | ||||||||||
| 35 | Individuals who perform in drag face potential ridicule or insult from straight people | ||||||||||
| 36 | Compared with other individuals in the gay community, the intelligence of drag queens is above average | ||||||||||
| 37 | Drag queens improve the image of the gay community to those outside the gay community | ||||||||||
| 38 | Drag queens gain much benefit from being drag queens | ||||||||||
| 39 | Some drag queens could be considered to be fakes | ||||||||||
| 40 | Individuals, while performing in drag, sexually arouse me | ||||||||||
| Other response items and associated questions | |||||||||||
| 41 | Individuals who perform in drag face potential ridicule or insult from members of the gay communitya | ||||||||||
| 41a | When do drag queens face the most ridicule or insult from members of the gay community?b | ||||||||||
| 42 | Drag queens are thought of as peculiar individuals in the gay communitya | ||||||||||
| 42a | In what way are drag queens thought of as peculiar individuals in the gay community?b | ||||||||||
| 43 | It is easy for a drag queen to damage their reputationa | ||||||||||
| 43a | In what ways could a drag queen damage their reputation?b | ||||||||||
| 44 | Is there such thing as a “fake” drag queen?c | ||||||||||
| 44a | What would make a “fake” drag queen?b | ||||||||||
| 45 | Do successful drag queens generally have a following of fans or other individuals?c | ||||||||||
| 45a | What kind of people usually comprise the group that follows a drag queen?b | ||||||||||
| 46 | Have you ever thought of performing in drag? If so, what are some of the reasons for why you considered performing?b | ||||||||||
| 47 | Why do you think most drag queens perform?b | ||||||||||
| 48 | What is your definition of a drag queen?b | ||||||||||
| 49 | Do you currently perform in drag at a nightclub, bar, or cabaret?c | ||||||||||
| 50 | How often do you perform in drag? | ||||||||||
| Never | Less than once a month | Once a month | 2–3 times a month | Once a week | 2–3 times a week | More than 3 times a week | |||||
| 51 | On average, how long do you believe it takes for a drag queen to become successful? | ||||||||||
| Less than 3 months | 3–6 Months | 6 Months to 1 year | 1–2 Years | 2–5 Years | 5–10 Years | More than 10 years | |||||
| 52 | The financial cost for a drag queen performing regularly is: | ||||||||||
| Not at all costly | A little costly | Very costly | Extremely costly | ||||||||
Note. The presentation of statements was randomized.
a7-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree). bFree response item. cYes/no response item.
Appendix B
Authors’ Note
Data will be made available upon the request.
Acknowledgment
We thank Daniel Benyshek, Andre Mentel, Matthew Martinez and Benjamin Purzycki for insightful and beneficial discussions, and two anonymous reviewers for their judicious comments.
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received financial support for the payment to the participants from the Department of Anthropology, University of Nevada, Las Vegas.