Introduction

MTV’s Catfish and NBC Dateline’s To Catch a Predator series illustrate society’s fascination with the toxic things that people do and say online; things that they might not do or say in a non-virtual context. While not all online actions are illegal, as in the case of Dateline’s predators, or secret, as with the Catfishers, the Internet has the ability to bring out rudeness and hatred that sometimes lives below the surface in a face-to-face setting. In the queer community, social networking profiles for men who have sex with men (MSM) provide one such outlet for toxicity. The first, and arguably most well known, mobile social network for MSM was Grindr, a social networking application that uses global positioning system (GPS) technology to connect MSM around the world with other MSM nearby. Grindr has inspired websites and groups aimed at tackling toxicity in the online MSM space. On the website Douchebags of Grindr, the rudest, most offensive, or downright hateful MSM-specific social networking profiles are screen captured and posted to the web with the goal of highlighting issues such as ageism, femmephobia, racism, and “arrogance” (among others). While countless Facebook groups have popped up to target these “Grindr douchebags,” other groups like “Freaks of Grindr” don’t so much expose people with problematic profiles, but poke fun at those who are considered ugly, gross, or somehow worthy of being mocked, thus extending the rudeness beyond the MSM-specific realm, taking it to Facebook and into mainstream society.

The present study sought to measure specific instances of toxic language use in MSM social networking profile advertisements. The aim of this project was to investigate the ways in which men self-present, as well as the factors by which they describe their ideal partner, on social networking applications for MSM. Because aesthetics play such a large role both in MSM culture and in online dating, the photographic elements of profiles were analyzed along with the textual self-descriptions. In examining these profiles, particular attention was paid to issues surrounding masculinity/femininity, race, age, and the gay male body.

Social networks for MSM have become increasingly popular in recent years, and with the advent of the smartphone came the GPS-based dating/sex application. There are no less than a dozen popular MSM-specific phone applications, ranging from the generic (e.g. Grindr, Jack’d, Hornet, etc.) to the niche (e.g. Scruff for hairier men, Growlr for chubbier men, MISTER for more mature men, etc.). This is not unpredictable, as gay men and women have been known to be more socially active online than their straight counterparts (“Gay and lesbian adults…” 2010; Mustanski et al. 2011). The current study builds upon some of the literature that discusses a strictly non-virtual gay community. It also extends a body of literature focused on gay male personal advertisements, much of which has not kept up with the swift pace of technological advancement. Very little is known about the communication patterns in profiles on mobile applications for MSM, and it is important to investigate whether findings regarding self-description and partner preferences in traditional personal advertisements extend to the mobile GPS-based social network. Examining how photographic elements impact the self-descriptions present in profiles, as well as how showing one’s face impacts what is said about potential partners, is a necessary next step in research on MSM’s online communication.

Literature Review

The Online Disinhibition Effect

Disinhibition, simply stated, is “any behavior that is characterized by an apparent reduction in concerns for self-presentation and the judgment of others.” (Joinson 1998, p. 4). Online, the dishinibition effect occurs when individuals say or do things on the Internet that they would not do or say in a non-computer-mediated setting (Suler 2004). This may occur in two ways: benign or toxic. Benign disinhibition involves the disclosure of very personal and thoughtful emotions, fears, or wishes, or even gratuitous acts of kindness or generosity (Suler 2004). On the other end of the spectrum, toxic disinhibition is anything but kind; this type of conduct might include rude language, hatred, cyberbullying or threats, crime, violence, or any other hidden ill behavior (Suler 2004). The toxic disinhibition effect may serve as an explanation for many of the unpleasant issues that have been highlighted on MSM-specific social networks—issues such as femmephobia, ageism, raceism, body shaming, and so on.

Suler (2004) identified six key features that might contribute to online disinhibition, including dissociative anonymity, invisibility, asynchronicity, solipsistic introjection, dissociative imagination, and the minimization of status and authority. Dissociative anonymity refers to the capability of cyberspace users to hide, or even modify, some or all of their non-virtual identities (Suler 2004). This anonymity might allow users to compartmentalize the offline and online aspects of the self and, therefore, to more easily avert responsibility for their online communication. Others have noted anonymity as a key element of MSM’s new media usage. For example, Gudelunas (2012) employed the uses and gratifications approach to examine gay males’ use of social networking. He found that, “one of the unique gratifications of gay-specific SNSs like Grindr is the sliding scale of anonymity provided” (p. 359). Invisibility, the next factor, refers to the fact that, in many online environments, people cannot be seen. Not only is their physical presence masked in text-based cyber avenues, but also virtual presence might even be safeguarded. In a study on toxic online disinhibition, Lapidot-Lefler and Barak (2012) use the term unidentifiability to discuss the ability of Internet users to remain relative unknown, even when some identifying personal details (e.g. real name) are disclosed. These authors list lack of personal information, lack of visibility, and lack of eye contact as the three contributors to online unidentifiability. Eye contact was found to be the strongest predictor of toxic online disinhibition (Lapidot-Lefler and Barak 2012).

While replies might occur in real-time in some cyber scenarios, most replies will not be instantaneous. As Suler (2004) notes, this asynchronicity creates disinhibition because it allows people to say or do something online and then wait until they are ready to experience the reaction, which might be never. Solipsistic introjection, or the merging of one’s mind and the mind of the online other, leads to a lack of inhibition because of the mind’s tendency to treat these others as characters in one’s own internal world (Suler 2004). The fifth factor, dissociative imagination occurs when people feel the characters they have manufactured exist in a distinct space, which allows for the “created” characters to be left in cyberspace once a person logs off (Suler 2004). The imagination is a safe place, and when online communication feels as though it is being experienced in such a space, it leads individuals to do or say things they would not say in non-virtual settings (Suler 2004). The last important factor affecting online disinhibition rests in the diminishment of status and authority. The absence of visual authority cues makes for a setting in which authority figures are often unperceived, even when they are present, and with no unified authority, the Internet operates as a space with minimized authority (Suler 2004). People act as if they have no one to which they must answer in part because of this perceived lack of authority figures.

One can imagine how at least some of these aspects—for example, dissociative anonymity—could contribute to online profiles that use language that people might not use as readily offline, particularly in the case of profiles that do not reveal a person’s face, name, or contact information. At the same time, online disinhibition allows for some men to find queer space that is safer and more readily accessible than the queer space offered to them in their everyday, non-virtual world. Therefore, there are both positive and negative aspects to viewing MSM-specific social networking through this particular theoretical lens, though I will now focus on some of the more problematic issues regarding MSM and cyberspace.

Personal Advertisements and MSM

Personal advertisements have long been an area of interest for scholars. As Lynn and Bolig (1985) noted, individuals place personal advertisements to attract other likeminded individuals, not to attract researchers. Therefore, the behavior exhibited and language used in personal ads is entirely representational of real life and has a high level of external validity. Personal advertisements also offer an abridged snapshot of self and partner preference descriptions (Gonzales and Meyers 1993). While Gonzales and Meyers (1993) were referring to traditional print (and paid) advertisements when making this assertion, we can apply this justification to the study of new media’s version of the personal advertisement: the online profile. It is on these profiles where we must focus our attention if we desire to learn about what MSM desire in a partner, and what they choose to highlight about themselves in their search for mates and dates, sex and love, and everything in between.

In the context of MSM, personal advertisements have been studied to investigate issues such as men’s style of loving (Lee 1976), to contrast gay male advertisements with those of heterosexuals (Laner and Kamel 1977), and to analyze HIV-status disclosure and related content (Hatala et al. 1998). Computer-mediated communication (CMC) is especially useful for MSM, as it creates unique gay space–space that may be unavailable, either geographically or culturally, to some men in an offline context. In an early social networking study, Shaw (1997) examined Internet Relay Chat users, determining that queer people were using the Internet in order to build community and meet new people away from the gay bar. This seems to have remained the case, though statistics on MSM Internet use for partner seeking varies. One study suggests that 85 % of MSM report some form of recent Internet use for the purpose of meeting other men (Rosser et al. 2011), while another puts the statistic at 35–45 % (Liau et al. 2006). In one sample of MSM in New York City, 55 % of participants reported logging onto web-based or mobile social networks for MSM at least daily (Grosskopf et al. 2014). Similarly, Lehmiller and Ioerger (2014) report an average of 3.03 daily log-ins for MSM-specific mobile application users, with an average engagement time of 11.75 min per visit.

Physical Characteristics and Gay Bodies

Gay men’s personal advertisements have traditionally been more focused on physical characteristics than those of lesbians, who tend to be more interested in emphasizing personality variables (Hatala and Prehodka 1996). Similarly, when Sergios and Cody (1986) examined the elements impacting dating and sex partner selection in MSM, they found that aesthetic variables were more important to gay men than internal characteristics. Even when MSM do not conform to society-wide definitions of what makes a man’s body attractive, much emphasis is often attributed to this alternate “sexy” body. In a study of gay “bears,” Gough and Flanders (2009) found that these larger, hairier men contrasted their community, “with a more judgemental ‘twink society’ defined by the policing of appearance” and the “perceived fixation on image and connotations of superficiality, perfectionism and falseness” (p. 244). Bears were understood in relation to their bodies, as well as their masculinity, while twinks were feminized for caring about their weight. Indeed, in spaces for gay men with “alternative” bodies, focus is squarely on the aesthetic. For example, Campbell’s (2004) investigation into Internet Relay Chat rooms for obese men, muscled men, and bears, highlighted the importance of online space for marginalized MSM bodies, but it also underscored the importance of the physical body in the online MSM space.

It is perhaps not surprising, then, that the literature on gay men and physical appearance makes linkages between body image and disordered eating (Duggan and McCreary 2004; Reilly and Rudd 2007; Rothblum et al. 1992). Gay men have been known to be more concerned with body shape and size than heterosexual men, possibly because of their media intake (including pornography) that includes more focus on the male body (Duggan and McCreary 2004). They also may engage in riskier eating behaviors simply because of the added minority stress of being visibly gay (Reilly and Rudd 2007). Hutson (2010) detailed the expectations of showing one’s body as a means of fitting in and of appearing “authentically” gay, evoking the concept of the “gay male gaze” (Wood 2004) to frame not only body dissatisfaction due to these expectations, but also identity dissatisfaction “based on the inability to establish a gay identity experienced as authentic” (p. 225).

Same-sex attracted males have also been found to exhibit a higher drive for muscularity than heterosexual men or women (Yelland and Tiggemann 2003). One study found that gay men have a high preference for low waist-to-chest ratios, indicating a robust idealization of upper-body muscularity (Swami and Tovee 2008). For gay men, affiliation with the queer community has been found to be a predictor of greater body dissatisfaction (Beren et al. 1995). However, while mere interactivity with the gay community has been associated with a higher drive for muscularity, perceived acceptance by this community is associated with lower drive for muscularity (Levesque and Vichesky 2006). Nonetheless, many men might have trouble living up to the pressures to be physically attractive (Siever 1994), youthful (Giles 1997), and fit (Drummond 2005), and this might cause them to question their identity as men who do not fit the gay ideal (Hutson 2010).

Sexuality, Racialised Language and Racial Stigma in MSM Online Space

Teunis (2007) studied gay men in San Francisco and found that they understood their sexuality in terms of their sexual position—positions that were linked to power as well as race. In this particular study, black men reported being put into a “top” position, even when their desire was for a more egalitarian or versatile setup. Contrastingly, gay Asian men have reported being stereotyped as subservient, as filling a “bottom” role in the bedroom and of wanting to please their partners in whatever way possible (Han et al. 2014a, b). For this reason, the current study examines disclosures of self-described sexual position, as well as partner sexual position preference through a racialised lens.

Gay men are more likely than their heterosexual counterparts to refer to their own racial background and their racial preference for partners in an online dating and sex-seeking context (Phua and Kaufman 2003; Plummer 2008). Because of the unique features of the Internet that create distance between users, these racialized references to potential partners are more common in virtual than non-virtual space (Plummer 2008). Particularly in White-dominated countries, gay men have been found to privilege the White race and the White body (Caluya 2006; Giwa and Greensmith 2012; Han 2008; Kendall and Martino 2006; Plummer 2008; Teunis 2007). Furthermore, when racial minorities are deemed attractive as mates in a queer culture, it is often in relation to stereotypes about specific racial and/or ethnic groups (Han 2008; Jackson 2000; Raj 2011; Ridge et al. 1999).

Therefore, it is interesting to note that, in an examination of Yahoo! personal advertisements, Phua and Kaufman (2003) found ethnical/racial minorities were more prone to requesting partners of a specific race than were White individuals. However, they also found that men seeking men were more likely to state a racial preference than their straight counterparts, and minority men tended to favor White partners (Phua and Kaufman 2003). In a more recent study, Callander et al. (2012) examined user profiles on Manhunt.net, a global dating site for MSM, in order to assess the ways that racialised language plays out online for the MSM community in Australia. They found that racialised language was being used for a variety of purposes, such as marketing of the self to others, negative discrimination (e.g. “No Asians.”), positive discrimination (e.g. “Mixed guys are sexy.”), or for commentary (Callander et al. 2012). Nineteen percent of analyzed profiles included some form of this race-focused text. It was also more common for people to use racialised language to refer to others rather than to refer to themselves or to race conceptually (Callander et al. 2012). White men were deemed the most likely to use racialised language to describe others, and Indian men were found to be the most likely to refer to their own selves in their profile text regarding race (Callander et al. 2012).

Payne (2007) suggests that gay men have created a racial hierarchy on dating websites, wherein White men are depicted as the top of the pyramid and Asian men as the bottom. He proposes that Asian men are constructed by White men as being outside of the regulatory acceptable image of gay male masculinity, and uses quotes from actual Gaydar.com profiles to make his case. Riggs (2013) has also examined anti-Asian sentiments on Gaydar, finding that discrimination is often framed as “just” a personal preference or an explanation of one’s type. Han et al. (2014a, b) propose that racial stigma against gay Asian men from within the queer community, “seems to be intimately tied to their perceived feminine traits,” traits that supposedly keep these men from reaching the masculine community norm. Indeed, perceptions and presentation of masculinity/femininity play a central role in the queer community, and particularly on MSM-specific social networks.

The Privileging of Masculinity in MSM Communities

Masculine practices and behaviors have been found to convey more symbolic value than those deemed to be feminine or womanly (Kimmel 1996; Messner 2000). Men are rewarded from an early age when they act tough or unemotional, as boys are told they should, and publicly sanctioned when they behave in ways that North American culture has labeled as feminine (Mirande 1997; Thorne 1993). These men who do not conform to normative gender standards are frequently labeled as female or gay (Pascoe 2007), and while feminine men are labeled as lesser than from outside of the queer community, many MSM have also been found to privilege masculinity over femininity (Bailey et al. 1997; Clarkson 2006; Harris 1997; Levine 1998). Often, these men promote masculinity as the ideal form of self-presentation (Clarkson 2006; Harris 1997).

Masculinity is also intrinsically linked with race and culture. For example, Pascoe (2007) found that the same tactics that White high school boys used to dub things as feminine or “gay” were the same strategies that Black high school boys used to affirm masculinity (e.g. dancing). Latino men may have a similarly calculated approach to the self-presentation of masculinity. In an ethnographic study of Latino MSM, Ocampo (2012) observed that self-presentation of masculinity was carefully cultivated. Latino men were particular about the way that they dressed, the way that they talked, and the social scenes in which they participated, but their specific preferences—for example, manicured eyebrows or lots of “bling”—would not be deemed as especially masculine in mainstream White culture (Pascoe 2007). Still, many Latino men were reluctant to identify as gay in the same direct way they identified their racial and ethnic background, and it was common for men in this sample to use homophobic slurs to assert and reaffirm their own masculinity, as well as to only seek partners who fit a stereotypical masculine mold.

In a study of anti-Asian sentiments on the Australian version of the website Gaydar, Riggs (2013) found that, along with “type” and “preference,” the words “feminine/fem/femmy/effeminate” were among the most used terms in MSM’s profile text. As West and Zimmerman (1987) note, masculinity in gender presentation operates as a way to adapt to the hierarchy that exists both between men and women, as well as between different groups of men. It is for this reason that we might see some groups utilizing more anti-feminine language than others. If a racial group is at the bottom of the hierarchy, there is no other racial group to devalue by feminization. Nonetheless, masculine privilege occurs within and across all racial groups and ethnicities in online queer space. When gay men were asked to define what made another gay man masculine/butch or feminine/femme, stereotypical personality and physical traits dominated the descriptions, and being able to pass as heterosexual or being “straight acting” was the second most repeated theme (Sánchez et al. 2009). This supports research on personal advertisements, in which gay men have been noted as extremely likely to both claim to possess, and/or to request masculine characteristics in their partners (Bailey et al. 1997; Laner and Kamel 1977; Lumby 1978). These include both stereotypically masculine appearance traits (e.g. muscularity) and activities (e.g. “sports fanatic”).

In an analysis of gay male and lesbian personals, Bailey et al. (1997) found that those who described themselves as either masculine or feminine were correspondingly more likely to explicitly seek masculine or feminine partners. Gay men have also been found to rate masculine gay others as more likeable than feminine ones (Skidmore et al. 2006). The preference for all things masculine transcends the boundary between virtual and non-virtual space; Gudelunas’ (2005) study of PlanetOut.com personal advertisements illuminated masculinity as a large concern in the free-form text areas of profiles, suggesting that men were looking for “straight-acting” or not “femme” partners.

Conceptions of masculinity have been linked with a number of problematic issues for MSM. For example, Hamilton and Mahalik (2009) found a connection between masculinity and risky sexual behaviors, and Halkitis et al. (2004) discovered a connection between masculinity and the use of anabolic steroids. The pressure to fit into traditional masculine roles has also been linked with internalized homonegativity (Szymanski and Carr 2008; Hamilton and Mahalik 2009), as has hypermasculinity in general (Estrada et al. 2011).

Accordingly, based on the literature, the following research questions were developed.

RQ1

How do men present themselves and their partner preferences in the textual and photographic elements of their profile?

RQ2

Will type of photographic self-presentation be related to the way men textually describe themselves, or the way men describe their partner preferences?

RQ3

Will there be differences by (a) race or (b) age in what type of photos, self-description, or partner preferences a user includes in his profile?

RQ4

Will there be differences by (a) muscularity or (b) weight in what type of photos, self-description, or partner preferences a user includes in his profile?

RQ5

Will men who mention their own (a) fitness level/body type or (b) masculinity/femininity describe their partner preferences differently than those who do not?

Method

Sampling

The sample included 300 profiles from the MSM-specific social networking application Jack’d. This social network was chosen for a number of reasons. First, it is a widely known application, used by many MSM across the world, and not targeted specifically to one demographic of men, as is the case with some similar applications (e.g. Scruff, GROWLr). Jack’d now has more than five million users in over 180 countries, with 80 % of them falling between the ages of 18 and 30 (PR Newsire 2014). Furthermore, Jack’d was chosen over other popular applications—namely Grindr—because of its randomized global feature. While Grindr only allows men to see who is located near them in physical proximity, Jack’d (and many other applications) offers a global feature that shows who is online around the world at any given moment. It also lets men search by specific location, opening up the boundaries that Grindr puts upon gay online space. Using Jack’d allowed for the use of a random sampling technique.

All included profiles were randomly selected from the list of online profiles at the time of data collection. Data collection occurred at six time points over the course of a weekend day (every 4 h, starting with 4 a.m. on Saturday and ending at 12 a.m. Sunday) in order to achieve a variety of profiles. A weekend day was chosen because of the possibility that traffic would be higher on a day when people are typically more social. In each instance, the first 50 online profiles were selected with a few filters. Profiles that were not written in English were excluded from the sample, as all coders were English-speaking and audience context was quite important to this project. In most instances, foreign profiles were at least mostly constructed in English and thus left in the sample. Profiles were also filtered to include only those advertisements that contained a photo, though photos could conceivably have been a black screen or non-personal image. Lastly, two repeated profiles that showed up at multiple time points were excluded on their second occurrence. Profiles were screen captured for the sake of accuracy.

Coder Training

Two doctoral students served as coders in the examination of these social networking profiles. One identified as a gay male, and the other as a heterosexual female. Both coders received training on terminology commonly encountered on MSM-specific social networks (e.g. bear, twink, masc, fem, poz, etc.). They were trained for approximately a week before coding the sample. Discrepancies were discussed and the codebook was refined numerous times. Reliabilities reported below were calculated using Krippendorff’s alpha, and an overlapping subset of the actual sample (12 %, n = 36).

Unit of Analyses and Operationalization of Variables

Because information could occur in a number of different places (e.g. the headline, in “interests,” or in “activities,” as well as other locations) and because both text and images were being examined, judgments were made at the profile level. A profile on Jack’d includes the headline, descriptive statistics (e.g. age, race, etc.), and the following open-ended and optional sections: about me, location, activities, interests, music, movies, books. I have grouped variables based on whether they relate to basic characteristics, photographic profile elements, textual self-description, or textual partner preference description.

Characteristics

Coders recorded a variety of demographic variables for the men depicted in the profiles in the sample. Some of these variables were commonly found included within the free-form text (e.g. location) and, thus, optional. Therefore, we do not have this information for all men. Other variables, such as race and age, are required components of the Jack’d profile. While this data is present for all users, we must keep in mind that Internet users have been known to be dishonest in their self-presentation and accept all information cautiously. Given the all-male nature of Jack’d, gender was not assessed, however, profile type (α = 1.0) gauged whether the profile belonged to a single male, a male/male couple, or other (which would include a woman, if she were to be present on the network). Age (α = 1.0), height (α = 1.0), and weight (α = 1.0) were all required profile components, though users were free to enter whatever numbers they wished, and there were at least a handful of obvious falsified entries (e.g. a reported height of 1′7″ or a reported weight of 63 lbs.). For the most part, entries seemed to be dependable at face value. Race (α = .96) was another required part of the profile, and users were made to choose between eight categories: White, Black, Latino, Asian, Middle Eastern, Pacific Islander, Mixed, or Other. Coders were instructed to also consult the text of the profile to determine race, particularly if the profile user checked “mixed” or “other.” Locations (α = .80) were optional information that some users decided to include, but many did not.

Photographic Content

Coders assessed a number of variables related to the profile’s visual components. They recorded whether the main profile photo showed full, partial, or no face (α = .80) and whether the main profile photo showed a man shirtless or not (α = 1.0). Shirtless was considered anything showing pectorals or nipples, abs or stomach, or all of the above. Overall type of main photo (e.g. face and body versus face versus cartoon) was coded for, but was not a reliable variable. Coders also noted whether profile users had uploaded secondary photos in addition to their main one (α = 1.0), and whether any of their photos showed them shirtless (α = .84). Attractiveness, weight, and muscularity (α = .77) were also measured based on visual content; however, the first two variables were not reliable and were dropped from the analysis.

Self-Descriptions

Coders recorded the masculinity/femininity, fitness level/body type, and sexual position self-descriptions contained within a user’s profile text, if such a description was presented. Age and race were not coded for in this regard, as these were required components of a profile. Coders looked for the overall presence of a masculinity/femininity self-description (α = 1.0), as well as the specific nature of that description; masculine (α = 1.0), feminine (α = 1.0), not masculine (α = 1.0), or not feminine (α = 1.0). Because each of these variables was coded separately, a man could have been coded, for example, as making a declaration about his masculinity as well as his lack of femininity. The men did not need to use the terms “masculine” or “feminine,” as coders were given specific instructions and examples on common MSM terms such as “straight-acting,” “masc” and “fem”, “butch” and “femme,” etc. Instances of fitness/body self-descriptions (α = .72) were assessed as any mention of the user’s own fitness level, interest in the gym or working out, or body type (e.g. “muscular”). Self-mentions of sexual position preference (α = .89) were coded for, as well as the specific position preference stated (α = 1.0).

Partner Preferences

Mentions of explicit partner preferences were also assessed, and coders examined preferences in terms of age, race, masculinity/femininity, and sexual position. Only the fitness/body type variable lacked reliability. Age preference (α = 1.0) was measured as explicitly mentioned or not mentioned, and both a preference for specific age ranges (α = 1.0) and an aversion to specific age ranges (α = 1.0) were accounted for. Racial preference (α = 1.0) was measured using a similar present/not present strategy. Coders also noted the exact type of preference/aversion based on racial group for Asian (α = 1.0), Black (α = 1.0), White (α = 1.0), Latino (α = 1.0), and other race (α = 1.0) men. The presence of a masculinity/femininity partner preference was recorded (α = 1.0), as well as whether the user explicitly mentioned wanting masculine (α = 1.0), feminine (α = 1.0), not masculine (α = 1.0), or not feminine (α = 1.0) mates. Finally, sexual position preference (α = 1.0) adjudged the sought sexual role users presented for their partners: desiring of top men, desiring of bottom men, desiring of top or versatile men, desiring of bottom or versatile men, or no stated sexual position preference.

Results

Demographics

Listed profile ages ranged from 18 to 99, though the latter age is an obvious attempt to avoid listing an accurate figure. The mean listed age was 31.59 (SD = 15.27). Twenty eight percent of profiles belonged to Black men (n = 83), which was closely followed by 26 % for White men (n = 78) and 26 % for Asian men (n = 77). Other listed categories of race included mixed (10 %, n = 31), Latino (9 %, n = 26), Middle Eastern (1 %, n = 4), and other (.3 %, n = 1). No men checked the racial category of Pacific Islander when creating their profile. The majority of men had no listed location (54 %, n = 162), as this was not a required textual element like race or age. Coders were instructed to look for location mentions in any part of the profile, revealing that 32 % of them hailed from North America (n = 96), 11 % from Asia (n = 32), 3 % from Europe (n = 8), and 1 % (n = 2) from Australia. A number of tests were run to answer the research questions, including Chi square analysis, bivariate correlations, regression analysis, and an independent samples t test.

Research Question 1

In terms of the variables of interest, men were found to explicitly classify themselves rather sparingly in the text of their profile. Nonetheless, there were instances of self-description that were noteworthy. While 6 % of men (n = 18) described their masculinity (or lack of femininity) in the text of their profile, not one man in the sample described himself as feminine, which provides support for the presence of at least some degree of femmephobia or pro-masculinity attitudes. A statistically significant Chi square goodness of fit test confirmed the differences in frequency, χ2(1) = 232.32, p < .001.

Despite the ability to upload photographs to convey fitness or muscularity, 19 % of men (n = 56) made note of their fitness level, body type, or gym interest in the text of their profiles. An even greater number of men had shirtless photos on their profile (31 %, n = 94). Seventy-eight percent of men (n = 234) showed either full face (21 %, n = 214) or partial face (7 %, n = 20) in their main profile photo, which is the photo that others see to determine whether they want to see more. That 22 % of men (n = 66) had faceless profile photos is not entirely surprising, given the sexual nature of MSM-specific applications that might inhibit the posting of identifiable photographs. It is also possible that the internet’s capacity to offer queer space to men who might not identify as gay or bisexual, or who may not be “out” in an offline context impacted this statistic. Self-presentation profile elements are presented in Table 1.

Table 1 Frequency of self-presentation elements in profile advertisements
Full size table

The second part of RQ1 asked about partner preferences in the text of the samples profiles. Six percent of men (n = 17) included an age preference for their potential partners in the text of their profile, more often listing a predilection for, rather than an aversion to, a particular age range. Only 3 % (n = 8) of men listed a racial preference for partners, which one can assume is because of the high number of racial minority profiles included in the sample. White men have been found to be more likely to use racialised language to describe others, versus non-white men who may use race-based language to describe the self (Callander et al. 2012). Interestingly, all of the racial preference mentions were positive (e.g. “I want Asian men”) versus negative (e.g. “No Asians”) in nature. Of course, by stating a preference for one type of race, users effectively exclude all other races, thus erasing any positivity from the equation.

Seven percent of men (n = 21) described a preference for masculine or feminine partners and, like with self-descriptions of masculinity and femininity, the results indicate an overwhelming privileging of masculinity. Only one man expressed interest in feminine partners, while 14 men explicitly stated that they wanted masculine partners, and eight men explicitly mentioned not wanting feminine partners. The results of a Chi square goodness of fit test were not significant. Data on partner preference elements of profile text are presented in Table 2.

Table 2 Frequency of partner preference elements in profile advertisements
Full size table

Research Question 2

As described above, 19 % of men (n = 56) included a self-description of their body, fitness level, or interest in the gym in the text of their profile. Men who described their body type, fitness level, or gym interests in the text of their profile were overrepresented in having uploaded shirtless photos, χ 2(1, N = 300) = 7.292, p = .007, V = .156, ASR = 2.7. It appears that men focused on representing their bodies in a specific way textually were invested in supporting that representation with visual documentation of fitness or muscularity. Shirtless photos were not associated with differences on any other self-description variable.

Description of one’s body type, fitness level, or gym interests was also associated with the presence of face in profile photos, χ 2(1, N = 300) = 5.709, p = .017, V = .138. Men showing full or partial face were underrepresented in providing fitness/body self-description (ASR = −2.4), and men showing no face in their main photo were overrepresented in this particular type of self-description (ASR = 2.4). In the absence of face in a photo, it is more likely for a particular body part—for instance, abs or chest—to be highlighted, which would coincide with muscularity or an interest in working out for many men.

It is also likely that the anonymity of being faceless online contributes to greater disinhibition (Suler 2004), which may lead some men to be more vocal about their particular tastes and preferences. For instance, having a faceless main profile photo was significantly correlated with self-descriptions of masculinity in the text of a profile, r(1) = .272, p < .001. Furthermore, faceless profiles were found to be overrepresented in mentions of a partner preference for masculinity/femininity, χ 2(1, N = 300) = 5.724, p = .017, V = .138, ASR = 2.4.

Research Question 3

Due to cell count size, only profiles belonging to White, Black, Latino, or Asian men could be used in Chi square analysis. While none of the partner preference variables showed differences by race, interesting patterns were found when race, profile photos and self-description text were examined. For example, racial differences were illuminated when investigating the presence of profile photos showing full or partial face, χ 2(3, N = 264) = 9.865, p = .020, V = .193. While White men were overrepresented in having photos showing their face (ASR = 2.5), Black men were underrepresented in having face-disclosing profiles (ASR = −2.6). No significant differences were found for Latino or Asian men’s profiles, and there were no racial differences wherein shirtless photos were concerned.

Compared with White, Asian, and Latino men, Black men’s profiles also had an overrepresentation of self-descriptions of masculinity, χ 2(3, N = 264) = 7.759, p = .051, V = .171, ASR = 2.8. This may perhaps be attributed to the relative lack of faces shown in Black men’s profiles, or perhaps it may be evidence of a larger cultural understanding of masculinity and homosexuality in the Black community.

Black men’s profiles were likewise found to have an overrepresentation of descriptions of their sexual position preference when compared to White, Latino, or Asian men, χ 2(3, N = 264) = 11.070, p = .011, V = .205, ASR = 2.6. Asian men, on the other hand, were significantly underrepresented in providing this type of self-disclosure (ASR = −3.0). This difference likely relates to the cultural variance between North America and other areas of the world.

Regression analysis was run to test differences in photos, self-description, and partner preference by age. No significant differences were found on any variables.

Research Question 4

Expectantly, users coded as visibly muscular were significantly overrepresented in having shirtless photos attached to their profile, χ 2(2, N = 300) = 35.702, p < .001, V = .345, ASR = 5.7. Interestingly, muscular men were also found to be overrepresented in having a main profile photo showing full or partial face, χ 2(2, N = 300) = 14.361, p = .001, V = .219, ASR = 3.8, which suggests that their focus on their bodies did not limit the inclusivity of facial features in the visual content of their profiles. Muscularity was not found to contribute to any other differences between profiles.

Reported weight was tested for using simple linear regression, and was found to be associated with both shirtless main photos, β = .134, t(299) = 2.336, p = .02, as well as shirtless photos in a either a main or secondary placement, β = .147, t(299) = 2.567, p = .011.

Research Question 5

As previously mentioned, there were significant differences in visual presentation for men who mentioned their fitness level or body type in their profile compared to those who did not. There also existed differences in their partner expectations. For instance, men who described their fitness level or body were significantly overrepresented (ASR = 3.2) in also having an explicit mention of partner sexual position preference (e.g. “Looking for tops”), χ 2(1, N = 300) = 9.946, p = .002, V = .182. It is interesting to note, however, that these men were no more likely to include a self-description of sexual position preference.

Men who mentioned their fitness level/interests or bodies in their profile text did not differ on mentions of masculinity as a partner preference, which is surprising given that self-described fitness level/interests or body type was correlated with mentions of self-masculinity, r(1) = .203, p < .001. There were, nevertheless, differences in the presence of age and race preferences.

Wherein partner age preference is concerned, men with self-descriptions regarding fitness or bodies were overrepresented in listing an age preference for others, χ 2(1, N = 300) = 6.014, p = .014, V = .142, ASR = 2.5. While cell counts were too small to run a Chi Square analysis on race preferences for partners, a bivariate correlation revealed that self-described fitness/body type was significantly correlated with racial partner preference, r(1) = .133, p = .021.

Only one partner preference variable was found to be correlated with self-described masculinity: descriptions of partner masculinity or femininity, t(17.469) = −2.434, p = .026, r = .253. Men who provided an explicit mention of their own masculinity were significantly more likely to explicitly mention a partner preference for masculinity.

Discussion

In a study of conceptions of gay community, LeBeau and Jellison (2009) found that most participants thought of the community as global in scope, which points to the importance of global-reaching social networks like Jack’d. Including profiles from across the world in the sample allowed for an investigation into the norms of a global gay culture. One limitation, of course, is the lack of precise data on the geographic origin of profiles. Since this was not a required field, it was not always clear where a profile user was located, which inhibited the ability to test for differences based on location.

LeBeau and Jellison’s (2009) study also pinpointed a number of perceived internal disadvantages of the gay community: the shallowness of other men in the community, having to deal with undesirable personality types, and ostracism or exclusion due to minority status (ethnicity, class, age, etc.). These issues seem to be even more prevalent online, due at least in part to the toxic dishinibition effect. The findings of the present study support the presence of troubling visual and textual components of MSM’s social networking profiles.

That not one man in 300 described himself as feminine in the text of his online profile coincides with the literature on masculinity and its importance within the gay male community, particularly in a dating or sexual context. This extends much of the literature on personal advertisements (e.g. Bailey et al. 1997; Laner and Kamel 1977; Lumby 1978) that has found a bias for men to self-describe as masculine over feminine. Thus, it seems that the technological advancements that allow for partners to be found in new ways has had little to no impact on the type of language used to self-describe gender performance. Despite the presence of photographs on mobile applications (versus in early print personals), for some men, much attention is still devoted to linguistically presenting the self as masculine.

One reason why anti-effeminacy might run rampant in online MSM spaces is due to the policing of masculinity that all men experience as products of a heterosexist and, often, homophobic culture (Anderson 2005). Male social spaces are known for their insistence on rigid masculine norms and hypermasculinity (Anderson 2005; Pascoe 2007). In fact, as Anderson notes (2005), “Once a social space is created for or claimed by men, the maintenance of that space is collectively policed by the social sanctions placed on men’s identities and behaviors as a whole” (p. 25). By upholding a masculine persona, MSM might simply be attempting to quash the stigma associated with being gay (Anderson 2005), even in a strictly queer space.

The tendency of men to engage in self-descriptions of muscularity is another important finding, and one that cannot be entirely separated from the findings about masculinity. Many researchers have noted the connection between gay male hypermasculinization and the muscular male body (e.g. Clarkson 2006; Halkitis 2001; Halkitis et al. 2004; Hutson 2010; Wood 2004). For example, in his analysis of StraightActing.com, Clarkson (2006) identified the dominant discourse as one about the aesthetic components of “straight-acting” masculinity. Masculine physicality and masculine traits were both seen as desirable; things such as physical power, large physiques, and other bodily elements one would associate with a muscular, fit male were deemed highly important in the assessment of gay male masculinity. In a study of telephone personals, Bartholome et al. (2000) found that more than three quarters of advertisements (81 %) included language about the body. Due to the lack of visual cues on telephone personal lines, this is not surprising, but given the ability of photographic representation on Jack’d, it is interesting that 19 % of men still felt the need to describe their muscularity or fitness level linguistically.

Thirty one percent of men were found to have uploaded a shirtless profile photo, possibly indicating a tendency for MSM to self-objectify. Future work should investigate self-objectification within members of this community, and within online spaces such as Jack’d, as well as more mainstream spaces like Facebook and Instagram. A simple instagram search for the hashtags “gay” or “instagay” provide anecdotal evidence that self-objectification is a widespread issue that spans beyond Jack’d and applications designed to facilitate MSM dating and sexual activity. This objectification of the self might be a product of what Campbell (2004) deems the gay male beauty myth, or “the hierarchies of beauty” that “are imposed on gay men in their understandings of their own bodies and the bodies they should desire” (p. 156). This beauty myth promotes the visual depiction of the hypermasculine, and thus hypermuscular, male body, which might contribute to a culture disposed to body oppression. Signorile (1997) calls this oppression “body fascism,” indicating “a rigid set of standards of physical beauty that pressures everyone within a particular group to conform to them” (pp. 27–28).

The large number of men who both uploaded shirtless photos as well as textually described their bodies or interest in fitness must be framed within our discussion of the literature on gay men, body image, and eating disorders. How online self-objectification couples with body image and disordered eating on social networks like Jack’d is a topic ripe for future inquiry. It would be interesting to note the reasons for which men self-promote in this manner, particularly as the presence of shirtless photos was correlated with this type of body-focused text. Furthermore, there tends to be far more half-naked self-presentation on MSM-only social networking applications than on mixed networks like Tinder or OkCupid, and researchers should examine the body presentation for heterosexuals in comparison to MSM.

Partner preferences were also explored in the first research question, and one significant finding was the relative lack of racial preferences. Racist language and racial stigma, in particular, has been a focus for a number of researchers examining MSM populations (e.g. Callander et al. 2012; Caluya 2006; Phua 2007; Phua and Kaufman 2003; Riggs 2013; Teunis 2007). The low levels of racist partner preferences are in line with previous quantitative work on the MSM application Gaydar (Riggs 2013). A study with more White MSM profiles might produce different results, however, this is not a definite. Scholars have noted that much other-directed racist language used by MSM in online spaces stems from White men (Callander et al. 2012). However, other research has found that White men are the least likely group of MSM to request a race, likely because they do not think about it (Phua and Kaufman 2003). This is a topic worthy of future inquiry and, in particular, scholars should examine the intersection between racial partner preferences and other partner preferences listed in profiles. Researchers may also think to compare the profiles on a number of different applications and websites to determine whether some are more hospitable to racial/ethnic difference, as we know very little about the differences in content posted on all of the various MSM-specific social networks.

It is also imperative that we consider the racial differences that were found in the current study. Black men were overrepresented in making self-descriptions about their masculinity, as well as in having faceless profile photos. They were also overrepresented in discussing their sexual positions, providing support for hypersexuality as a defining feature of hypermasculinity. While this all may point to online dishinibition, it might also be evidence of the cultural conception of masculinity and queerness in the black community. Hunter (2010) found that not all Black gay men identify with the concept of a gay identity, which might significantly impact how they view feminine gay men and effeminacy in general, as well as their degree of outness as individuals with same-sex desires. If a man is less likely to be out or to identify as gay, one would assume he would be less likely to show his face on his online profile on a network clearly designed for MSM. Many other scholars have noted the tendency of Black MSM to reject the gay label (e.g. Cochran et al. 2004; Han et al. 2014a, b; Lapinski et al. 2010; Martinez and Sullivan 1998)

The findings on Asian men are also worthy of discussion, as this racial group was underrepresented in disclosing their sexual position preference. We must take into account that a large number of profiles of Asian men were from men who live in countries like Singapore, where male homosexuality is still illegal. Thus, their profiles may be “cleaner” or devoid of sexual content in case they were ever to be charged with a crime or called out for being on a social network for MSM in the first place. By not including sexual content, these men could effectively claim to simply be looking for friends. Future research should compare and contrast the profiles of Asian-American men and Asian men who live overseas. As previously stated, this is a large limitation of the current work; as location was not a required field, the specific location of most profile users was unknown. It was therefore impossible to test for differences between Asian American men and Asian men residing in Asia.

That men who showed their face were underrepresented in providing fitness/body self-description, and men showing no face in their main photo were overrepresented in this particular type of self-description is interesting. Perhaps the posting a face photograph exerts a certain level of power that serves as capital in this particular online space. As Mowlabocus (2010) notes, the face photograph can be seen as a response to gay men’s historic invisibility, and a commitment to being present and authentic in the online space. By affirming this commitment to a visible gay or bisexual identity, perhaps men are given a form of capital, and if so, men who do not post a facially-identifiable photo might need to compensate for the lack of capital by focusing on the body. They may do so visually, by sexualizing their body in their photographs, or linguistically, by describing their bodies.

The anonymity afforded by remaining faceless online likely contributes to greater disinhibition (Suler 2004). This may indeed lead some men to be more vocal about their desires in a partner. That having a faceless main profile photo was significantly correlated with self-descriptions of masculinity, as well as mentions of a partner preference for masculinity/femininity, suggests that the men who do not post their face are more invested in maintaining a hegemonic masculine ideal, both for themselves and for other MSM. This may be correlated with internalized homonegativity, a lack of identification with a gay or bisexual identity, and/or other psychological constructs (e.g. masculinity consciousness), and future work should employ survey methods to make these connections.

We might hypothesize that perhaps some “faceless” men might be in opposite or same-sex offline relationships and, thus, remain faceless for the sake of anonymity. Men might also lack a face photo because of their career, their sexual interests, or for a multitude of other reasons. We know very little about why men make the choices they do when it comes to profile photo selection, and this would make a fascinating study in itself. In a content analysis, it is impossible to know the motivations behind remaining faceless online, and future work should use alternative methods to engage this question. As well as issues of privacy, there is the issue of public perception of the MSM-specific mobile application, and of the associated psychological and sociological constructs that might lead to a culture of “participatory reluctance,” wherein men are primarily dismissive of MSM-specific online media and their participation as users of this media (Cassidy 2013).

In a number of instances, self-description variables (e.g. self-descriptions of fitness/body type and of masculinity) were correlated, and the same was true for some self-descriptions and partner preference variables (e.g. self-description of fitness/body type and age preference for partners). Future research might comprehensively examine the co-occurrences of, and interactions between, different types of self-description and partner preference variables. Further survey and experimental work is needed on MSM-specific social networking as well, as much of the literature tends to be qualitative in nature. HIV-stigmatizing language, slut shaming language (including language used to discuss monogamy, open relationships, polyamory, kink, etc.), and other semantic toxicity are other areas that should be explored.

The present study contributes to our understandings of how gay men approach online interaction, specifically on GPS-based phone applications, a technology that seems to be here to stay. It also advances our knowledge about self-presentation and partner preferences in MSM. In particular, it opens up discussion about how men frame their own age, race, masculinity/femininity, and bodies, as well as the age, race, masculinity, and bodies of others they do or do not find desirable. Lastly, this study investigated the visual choices that men make in terms of online profile photos, paying fastidious attention to the presentation of the face and the presentation of the semi-clothed body.