Gay / Bisexual Male Youth Suicide Problems The Original San Diego Paper Colorado Paper ('97) - - Banff Paper ('95) - - 1997-98 Papers

An updated interim paper prepared in support of a Poster Presentaton with the same name at the 2004 CASP (Canadian Association For Suicide Prevention) Conference held in Edmonton, Alberta. Original Paper.

Anti-Gay Harassment: Masculinity, the Feminine "Other," and Elevated Rates Of Suicidality

There are media and research reports that GLBT (gay, lesbian, bisexual and transgender) individuals in society and public schools experience high levels of anti-gay harassment, abuses, and violence (e.g., Herek and Berrill, 1992; Hershberger and D'Augelli, 1995; Human Rights Watch, 2001). These reports may lead people to believe that only GLBT people experience anti-gay harassment however, this assumption would be incorrect. Adolescents targeted for anti-gay harassment in public schools, for example, are mostly assumed to be homosexual. This is evident from many GLB youth descriptions about their lives in public schools where they felt alone and fearful of being hated if others were ever to discover their secret, and had no one to talk to about their homosexual feelings, desires and problems. They usually did not know of peers who were gay or lesbian in their school or neighbourhoods, although a few individuals may have been suspected, especially if they manifested gender nonconformable attributes. As reported in a Canadian Public Health Association study of youth living in large to small cities, "GLB youth almost universally experience a sense of isolation… [that happens to be] the most relentless feature in the lives of most gay, lesbian and bisexual youth. And the isolation is more profound than simply social and physical: it is also emotional and cognitive" (CPHA, 1998: 4-5). The fact that anti-gay harassment is generally directed at adolescents assumed to be gay, lesbian, or bisexual is highlighted in the 1995 Seattle Youth Risk Behavior Survey data (Table 12).
 
 

The 1995 Seattle YRBS results indicate that 5.0 percent of adolescent males (185/3685) are targeted for anti-gay harassment in schools or on the way to school and that GB (gay and bisexual) males account for only 19.4% of all males targeted (36/185). This represents 28.6 percent of all GB males (36/126). "Heterosexual" adolescent males, and those "Unsure" about their sexual orientation account for 67.6 (125/185) and 13.0 (24/185) percent of males reporting anti-gay harassment respectively, thus representing 80.6 percent of males targeted for anti-gay harassment. This is fairly conclusive evidence that the vast majority of males (80.6%) subjected to anti-gay harassment are not "known" to be homosexual, but are "presumed" to be so. The same would likely apply for most GB adolescent males similarly harassed. The attempted suicide rate (12 month period) for heterosexual males targeted for anti-gay harassment (16.0%) is almost equal to that of GB males targeted (19.4%) or not targeted (20.0%) for anti-gay harassment, compared to a much lower rate (4.3%) for heterosexual males not targeted for such harassment. These estimates suggest that anti-gay harassment / abuse is associated with significant distress for these males, resulting in an almost quadrupled attempted suicide rate, compared to heterosexual males not assumed to be gay and not harassed for this reason (RR, 95%CI = 2.4<3.7<5.7).

Some research reports state that the experience of harassment/victimisation is not implicated in higher levels of suicidality for GLB youth (Hershberger and D'Augelli, 1995), but others report there is an association (Nicholas and Howard, 1998; 2001). The Seattle results support the former proposition given that antigay harassment does not appear to be associated with GB males attempting suicide since those targeted or not targeted for harassment have similar suicide attempt rates: 19.4 and 20.0 percent, respectively. This conclusion, however, could be an "assumption" based on a lack of knowledge concerning "all" the factors that would/could be implicated in GB male youth suicidality. For example, varied sources of information (e.g. GB youth themselves and writings on the Internet) have led to the suspicion that closeted GB youth could be negatively affected when they are witnessing others targeted for anti-gay abuse. Some of these youth would also experience much guilt and self-reprimands given that they feared intervening because others might then suspect one's own homosexual orientation that ostracism and abuses would likely result. Such behaviour, however, could easily lead one to think that he is, in fact, living up to the "wimp" gay stereotype, including not having courage or integrity. This suspicion is documented in a Master's thesis study (N = 329) on mostly white South African gay and bisexual males (Theuninck, 2000). A request was made to Anthony Theuninck for comments on the Seattle results in relation to his study that had explored factors associated with gay males attempting suicide for reasons related to their homosexuality.

"What could account for these vulnerabilities to suicide? It must be remembered that many factors are associated with youth suicidality, of which gay-related stressors may form an important part, especially for males. In the case of male youth exposed to anti-gay abuse, the gay-related stressor is clear, namely an external source of threat and harm that is experienced because one is associated with homosexuality (accurately or not). But this leaves the question as to why gay youth who have not been exposed to anti-gay harassment show a similar suicide attempt incidences.

A possible explanation may lie in findings from my research in which a correlational matrix of gay-related stressors caused two covariance groups to emerge that can be labelled as internal and external stressors (Theuninck 2000). External stressors include events that occur largely independent of a person's perception of them such as physical and verbal violation. Internal stressors included internalised homophobia and the perception of stigma (i.e. perceiving society as discriminating against gay people). These stressors rely on an internalisation of societal norms and threats. Through internalised homophobia the person adopts the beliefs that homosexuality is not legitimate, that it is a sickness, moral weakness, defect, or some such deformation of self. This may lead to intense self-loathing related to having same-sex desires.

Through stigma the person learns that society is hostile to homosexuality and poses a threat to those so identified. Such a stressor creates an overall context of fear in which a person may fear his/her sexuality will be known, or may fear and live in trepidation of the ever present possibility of being attacked or discriminated against in the course of daily life. Such stressors could account for suicidal tendencies in people identified as homosexual or bisexual but have not been exposed to overt external stressors such as anti-gay harassment.

Even in cases where anti-gay harassment is experienced, it could be argued that the more pernicious stressor is the internal stressors, and more specifically internalised homophobia. In my research I examined the regression relationship of minority stressors and traumatic stress symptoms with gay-related suicidality (i.e. thinking about or attempting to commit suicide because one has homosexual desires). The stressors that associated most strongly with gay-related suicidality were internalised homophobia, followed by witnessed gay victimisation and negative fundamental beliefs. By negative fundamental beliefs is meant those beliefs that are essential to maintaining the coherence of self as defined by the works of McCann and Pearlman (1990).

Where homophobia has become internalised, where a person has colluded in victimisation of other gays, by taking part or standing by, in these cases it seems that the stress of the "gay scourge" becomes more pernicious. The evil is within the individual. The person's sense of self comes under direct assault by himself. His fundamental beliefs about being good, in control, safe in the world, and trustworthy start to untangle, for he is his worst enemy. His gayness becomes the source of torture, rather than seeing the external world as torturous of his sexuality. Under these circumstances he is more likely to have thoughts about killing himself for having gay urges and also attempt suicide." (Theuninck, 2002)

1. Self-identified as gay or bisexual males who are or are not targeted for anti-gay harassment, the latter likely being in the closet and also suffering from feelings of isolation (CPHA, 1998). They would also be witnessing anti-gay harassment on a day to day basis with likely negative effects (Theuninck, 2000, 2002). Their 7.6 times greater risk for experiencing anti-gay harassment compared to heterosexual males (RR, 95%CI = 5.5<7.6<10.6; c²= 169.2, p < .0000) suggests that they may have the highest level of gender nonconformity in the three "sexual orientation" categories. As a group, they are at significantly greater risk for a suicide attempt compared heterosexual males: RR, 95%CI = 2.8<4.1<6.1; c²= 54.5, p < .0000. They are also at much greater risk for a suicide attempt associated with receiving medical attention, compared to heterosexual males: RR, 95%CI = 4.0<7.0<12.2; c²= 57.6, p < .0000.

2. Males who are "unsure" about their sexual orientation, as in not claiming to be heterosexual, likely includes males who suspect that they are gay or bisexual and are troubled by this possibility. Their 3 times greater risk for experiencing anti-gay harassment compared to heterosexual males (RR, 95%CI = 1.9<3.0<4.5; c²= 27.7, p < .0000) suggests that higher level of gender nonconformity may exist for these males, compared to heterosexual males. In this group, individuals targeted for anti-gay harassment are at significantly higher risk for a suicide attempt than their non-targeted counterparts: RR, 95%CI = 1.3<2.6<5.6; c²= 6.2, p = .013. They are also, as a group, at significantly higher risk for a suicide attempt than heterosexual males: RR, 95%CI = 1.9<2.7<4.0; c²= 27.4, p < .0000.

3. Males who report being heterosexual identified but, because they are suspected / believed to be homosexual, they are being harassed / abused accordingly. They are at significantly higher risk for a suicide attempt than their non-targeted counterparts: RR, 95%CI = 2.3<3.7<5.7; c²= 35.8, p < .0000. The same applies for suicide attempt(s) associated with receiving medical care: RR, 95%CI = 2.2<4.7<9.7; c²= 19.8, p < .0000.

The Seattle study raised several important questions about the homosexuality factor: Why are so many heterosexual adolescents reporting that they are assumed to be homosexual and are being harassed accordingly? What kind of criteria are adolescents using to make such a determination? These questions have received minimal research attention as the answer for most is "obvious." Therefore its seems that researchers have not wanted to know, or have confirmed by quantitative studies, exactly how adolescents or adults arrive at the suspicion/assumption that an individual is homosexual. Only one exception was located. In Lyon, France, Daniel Welzer-Lang and Pierre Dutey did ask about 500 adults, via a questionnaire, if they had ever encountered individuals in public places that they suspected to be homosexual and, if so, what was the criteria used to make such a determination. Ninety-five percent of adults reported having made such determination and 90 percent of them assumed that the question only applied for males, but the question had been gender non-specific. The most common criteria used to suspect that a male was homosexual were manifestations of feminisation: how an individual dressed, gestures, and the way the individual talked They also noted that these homosexual male stereotypes have been the rule in the media in France (Welzer-Lang, 1994: 40-1).

It would therefore seem that "the obvious" with respect to homosexuality remains unstudied. Most researcher of school "bullying" (the term most commonly used to speak about verbal, emotional, and/or physical abuse in schools) also have a history of not investigating socially constructed attributes, such as homophobia and sexism, that may be implicated in the bullying experienced by children and adolescents. Percy-Smith and Matthews (2001) speak to this:

"In contemporary social construction, bullying is increasingly seen as a problem. Yet a paradox exists in that, for example, the systematic abuse of power has, throughout history, characterised many processes of social interaction, for instance in the workplace by bosses, by teachers in schools or by parents in the home, and continues to do so in some cases. To this extent, bullying is a culturally and historically bound concept varying according to individual as well as dominant social assumptions of acceptable and unacceptable behaviour… Swain (1998) suggests that any act of bullying involves dominance and subordination. Bullying can therefore occur whenever an unequal power differential, perceived or real, exists. This could be according to physical size, intellect, age, ability, sexuality, race, class, gender, culture, knowledge, status, force of personality, or brute force… These perspective are not designed to provide a definitive statement about what constitutes bullying, but rather seek to underline the dilemma of understanding and responding to bullying as a social phenomenon" (p. 51).

"However, anti-racist education in Britain has not been well served by researchers into bullying in schools. Loach and Bloor (1995) even argue that anti-bullying work is itself racist - including racism, sexism, homophobia, and other 'isms' - can be hidden. To focus on the overt aspects of bullying behaviour, they claim, is to ignore the background dynamics of personal and social relationships, which means that the behaviour can never really be understood. How valid is this criticism?" (Eslea and Mukhtar, 2000: 208)

Hawker and Boulton also noted a number of methodological problems in the studies of school bullying. Surprisingly with respect to suicide problems associated with this kind of victimisation (e.g, Rigby, 1996), they state: "Data concerning unhappiness and suicidal ideation are not included in Table 3, as there was an abundance of studies which used well-validated measures of depression…" (447). This comment indicates that many bully interested researchers somehow believe that measures of depression will suffice when exploring the negative effects of peer victimisation (possibly because depression is often said to be associated with suicidality) and that suicidality measures, such as attempting suicide can therefore be ignored. Fortunately, the authors did attribute an importance to the "variables such as disability, age, gender, and sexual orientation," but only because "it would be valuable to investigate [their] moderating effects on victims' adjustment." A need for "studies of children from cultural backgrounds that are not Scandinavian or English-speaking" was mentioned, but only because most of the bullying research has included only these groups (453). Although such recommendations advocate for an expansion of the existing knowledge about school bullying, they likely fall short with respect to venturing where one must go to better understand these problems.

To date, researchers have failed to explore how peer related harassment / bullying / victimisation is related to one's minority ethnic status. They have also not explored sexual orientation issues such as the highly stigmatised homosexuality that is implied when male children and adolescents are manifesting some degree of femininity. Unwanted "sexual touching" that may occur in schools and elsewhere is certainly a form of harassment and victimisation, but mainstream researchers of bullying in childhood and adolescence have generally ignored this issue, and the same applies for "forced sexual intercourse" that is certainly a form of victimisation. Is it possible that such serious sexual assaults, if inflicted on a male by another male, could also come under the rubric "targeted for anti-gay victimisation" that has been so prevalent in schools and elsewhere? A recent Internet study indicates that, for self-identified American gay and bisexual males ranging in age from 13 to 18 years, "forced sexual acts" are in many cases inflicted on them because, as it was specified in the questionnaire, "you are, or were thought to be, queer?" Out of the 1,763 gay and bisexual adolescents who responded to this question, 7.3 percent replied in the affirmative to having been harassed in this way, but many more have been harassed in other ways (Table 13). "Forced sexual intercourse" would be a subset of "forced sexual acts," and it would be the equivalent of rape. As the results listed in Table 13, indicate, self-identified gay and bisexual youth experience high levels of harassment in their lives ranging from verbal abuse (63.8%) to being physical assaulted (11.1%) and also being threatened with a weapon (4.7%), and most of these are associated with increased levels of suicidality.
 
 

The harassment of gay and bisexual males ocurs in varied forms and, as the result of the 1995 Seatlle Youth Risk Behavior Survey indicates (Table 12), they are not he only adolescent males subjected to anti-gay harassment. Of the 185 males reporting anti-gay harassment (5.2% of males), 19.4 percent (36/185) were gay/bisexual identified, 13.0% (24/185) were unsure about their sexual orientation, and about two-thirds (67.6%: 125/185) were heterosexual identified. The incidence for attempting suicide by males subjected to anti-gay harassment and those not subjected to this harassment was 18.4% and 5.3%, producing an Odds Ratio of 4.0, 95% CI = 2.7-6.0, p = .0000. This Odd Ratio for "attempted suicide" is similar to the one produced from an analyis of all males in the 1999 Oregon Youth Risk Behavior data set by the authors (Table 14): 4.0<5.1<6.5, p < .000.

In 2001, with the objective of exploring some forms of harassment generally ignored by many researchers, and especially with the objective of exploring associations with suicide behaviors that generally represent high level of distress, we acquired the 1999 Oregon Youth Risk Behavior data set. basic The basic statistical data related to ant-/gay harassment is presented on the basis of race / ethnicity in Table 14 and Table 15. The analysis was accomplished by using the SPSS 10 statistical program and recoding the six race / ethnicity variables (self-identification by study subjects) to produce four (4) combined and seven (7) more discrete "race / ethnic" categories for analysis. These groups are:
 
 

For these 11 groups of males, results were generated for the relationships between school related "Anti-Gay Harassment" (harassment because one is presumed to be gay or bisexual) and five "suicidality" variables reported to have occurred over the past 12 months:
 
 

The varied statistical associations between anti-gay harassment and the five suicidality variables (Table 14, Table 15) indicates that adolescent males are at risk for suicidality in association with anti-gay harassment, and that this risk generally increases for the more serious forms of suicidality. Variations in suicidality incidences and odd ratios also lead to the recommendation that researchers should not produce statistical results on the basis of ambiguous categories such as "All Males" or "Males Of Colour." It must also be recognised that a category like "Asian" includes very diverse groups of people and that they may not all would respond to certain life situations in the same way. This would apply for French and English Canadians with respect to some attributes being investigated. It also applies for North American people of Japanese, Chinese, Korean, South Asian, and Southeast Asian descent, noting here that these categories also include varied cultures, often with different religions, and even different views on homosexuality. The human categories used in our analyses are therefore lacking in many ways, but the results related to the listed "race / ethnic" categories, as solicited in Youth Risk Behavior Surveys, may be better than results derived from collapsing the data into an "All Males" category as it has been done in most published studies based on YRBS data sets. These data set are smaller (e.g Massachusetts YRBS has about 2,000 males, compared to 11,316 in the Oregon YRBS data set), thus severaly limiting what can be said about significant suicidality associations for adolescent males in "Ethnic / Race" minority groups. Results presented in Table 14 and Table 15 indicate that low counts preclude statistical analyses, such as Logistic Regression, within most "Ethnic / Race" minority categories in the Oregon data set, and more so with respect to the serious forms of suicidality. This problem, however, is 5-times worse in data sets containing only 2,000 males, thus suggesting that racism be apply on the part of the designers of such studies studies. That is, they are essentially manifestatiing a form of racism that does not recognize the possibility that people of colour experience life situations different than those of white people, and maybe that all people of colour are the same, when major differences may exist and should be studied.

Our work on the 1999 Oregon YRBS data set indicates that, when all males are analysed as a group, important factors existing within a self-identified group (that may be quite different that factors existing in other groups) are often washed out, thus greatly limiting the value of generated statistical associations. The same applies for categories such as "Of Colour" or "Of colour - One Race," the latter meaning that adolescent males studied reported a self-identification with only one "race / ethnic" group. Of great significance are the results for adolescent males identifying as "Multi-Race" (N = 282, Table 14) because, in the "Of Colour" category (N = 2,048) where these males account for 13.4 percent of the group (282/2,048), they account for an significant overrepresentation of  suicide attempters: 30% (30/100). For the more serious forms of suicidality (multiple suicide attempts and/or suicicide attempt(s) associated with having received medical attention), however, their overrepresentation is larger: 36.4 percent (20/55). On the other hand, "Latino / Hispanic males make up 33.9 percent of the "Of Colour" group (694/2,048, Table 14, Table 15) but they only account for 16.4 percent (9/55) of the suicide attempters. With respect to the "Anti-Gay Harassment" variable, the differences in associated suicidality may be even greater as illustrated by "Multi-Race" males accounting for 51.2 percent of the more serious suicide attempters reporting "Anti-Gay Harassment," while "Latino / Hispanic" males account for 0.0 percent of these suicide attempters. Therefore, combining such males groups in statistical analyses to produce basic incidences for varied suicidality variables or suicidality incidences associated with negative life experiences (e.g. experiencing anti-gay harassment because one is assumed to be gay or bisexual), to generating predictors via multivariate analysis could become an exercise bordering on the ridiculous, especially with respect to the interpretive value of some results. That is, if groups are to be combined for analysis, this should be done for a reason. Furthermore, the groups combined for analyses should also have similar incidence rates for the dependent variable and somewhat similar sizes so that the counts of one group (maybe having 500 invividuals, with a suicide attempt incidence of 2%) are not overwhelmed by the counts of another group (maybe having 10,000 individuals and a suicide attempt incidence of 6 percent). Furthermore, in procedures such as binary logistic regression analyses with combined groups, it would be important to know the group origins for the counts producing the predictors given that a predictor may not apply to one or more of the group in the statistical analysis. This is illustrated in Table 16 where the results of a Binary Logistic Regression analysis results for three groups of males are presented.
 
 

Native Americans (one race, N = 284), Native Americans (Multi-Race, N = 164), and other Multi-Race males (N = 210) were combined for the Logistic Regression Analysis because they had the highest "attempted suicide" incidence (6.2%, 7.5%, 6.6%, respectively) of all male groups (Whites: 2.8%, Others averaging 3.1%) and because they are related via the Native American, Multi-Race group. A major result of the analysis was experiencing "anti-gay harassment" (applies for 47 percent of males - 8/17 - who reported multiple suicide attempts and/or attempt(s) associated with receiving medical care) was a significant predictor, and that the majority of males in this category (75%: 6/8) were also reporting being "slapped/hit at home by an adult." This variable, however, did not remain significant in the regression mostly because it was being dominated by the "anti-gay harassment" variable, but it would have been a significant predictor if the information about "anti-gay harassment" had not been solicited and this variable was not on the regression analysis. In other words the two variables covary and it is also likely that the same factor causing boys to be subjected to "anti-gay harssment" in school (most likely gender nonconformity and related suspicions that one is homosexual) is also resulting in physical assaults at home. The Corliss et al (2002) research findings support this hypothesis given that, from an American national sample, gay/bisexual men were most at risk, compared to heterosexual males, for having experienced major physical maltreatment by their mothers and fathers (ORs = 3.6, 3.8, respectively: multivariate logistic regression analysis), and at lesser risk for "any physical maltreatment" and "emotional maltreatment." Such research result as well as other cited by Corliss et al (2002), combined with the results noted for "at risk" Native American and Multi-Race adolescent males assumed to be gay in Oregon high schools, and the fact that most of them are also being assaulted at home, suggest that homosexuality issues are intimately related with adolescent male suicidality, and especially with the most serious forms of suicidality. That is, "anti-gay harassment at school, by itself, but with other likel similar negative experiences also applying, accounts for almost 50 percent of the Native American and Multi-Racial adolescent males who attempted suicide multiple times in a one year period and/or received medical care as the result of one or more suicide attempts.

Little has been reported in the research literature on the relationship of harassment, including anti-gay harassment, and the suicide problems of adolescent males that are also strongly associated with the often unsolicited "gay/bisexual" identification factor and moderately associated with the "unsure of one's sexual orientation" factor. These life situations are all implicated in adolescent males suicidality as suggested in the results of the 1995 Seattle Youth Risk Behavior Survey (Table 11) and in the 1995 Massachusetts' Youth Risk Behavior Survey (Note 6). Issues related to male homosexuality presented throughout this paper, and especially those related to stigmatisation and related abuses, serve to further understand adolescent male suicidality. This applies for adolescent males who have recognised same-sex sexual desires and are reporting themselves to be gay or bisexual and/or to be engaging in same-sex sexual activity, and it also applies for adolescent males "assumed" to be gay and are abused accordingly. They are at risk, especially because these abuses are occurring in environments where homonegativity / homophobia have been the norm and are acceptable. To this day, such attitudes and related abuses rarely result in reprimands from anyone and therein lies part of the problem, but there is much more to this story.

Homophobia is said to be "a weapon of sexism" (Pharr, 1993) and this would mean that most males are still growing and actively working to maintain their supremacy status vis-à-vis females. How most boys and men continue to really feel about females is usually not expressed directly, but their inner feelings become apparent when males are responding to a feminine male, or a male deemed to be feminine. He will be laughed at or insulted, and these responses are always rooted in the "learned" traditional perception that femininity is a degraded status. The link between homophobia / homosexuality and sexism is replicated whenever a male is assumed to be homosexual because he is feminine to any significant degree, and this belief "is culturally supported by a powerful cultural stereotype" (Frommer, 2000: 194; Connell, 1992: 748). The assumption reflects the long time belief that all males should only be sexually attracted to females and vice versa. This means that everyone is to be 100 percent heterosexual and that only heterosexual people should exist, which is the standard definition of heterosexism (Stevenson and Medler, 1995: 1). It is also embodied in the expression "heteronormativity" used to describe school environments by Morino (2000). According to Epstein and Johnson (1994), heterosexism is the presumption of heterosexuality and it underpins the problem of homophobia (Beckett and Denborough, 1995: 112). Homophobia is strategically deployed by adolescent males to keep other males in line (Plummer, 1999, 2001; Morino, 2000: 227), and these abuses are often carried out by males acting collectively in "the policing of masculinities" that is intimately related to boys having acquired highly negative attitudes vis-à-vis females (Marino, 2000). The "policing of masculinities" is hegemonic masculinity in action and all males manifesting feminine attributes (and assumed to be homosexual) will be targeted for exclusion and/or abuse because, as noted by Connell (1992: 748), we live in "a patriarchal society in which hegemonic masculinity is defined as exclusively heterosexual." Many of these hegemonic masculine boys will also become the young adults males described in Fiona Hart's 1990 study: "The construction of masculinity in men's friendships: misogyny, heterosexism and homophobia." Others will be part of the athletic subculture where "women… [and] men who [do] not display traditional hegemonic characteristics like aggression, domination and loud banter" will be excluded (Harvey, 1999: 105). Nayak and Kehily (1996: 214) describe the association between hegemonic masculinity, misogyny and homophobia:

"Lees (1993, pp. 89-90) remarks 'it is an insult for a boy to be called a 'woman' because being similar to girls is also to be associated with a lower status group'. The connection between hegemonic masculinities and homophobia can be seen when we consider that many of the characteristics ascribed to gay men are applied to women. In this sense homophobia and misogyny can overlap where one is 'spoken' through the other. An example of this occurs in Lehtonen's (1995) study of young people's sexual identities in Finnish schools where he remarks on the use of the term 'vitun homo' meaning 'fucking queer' which is made up of a combination of the words 'cunt' and 'gay'. Homophobia is then a means of consolidating sexuality and gender through the traducing of femininity, and its association with homosexuality… Young men may worry about being gay, and being called gay, in part due to the intense hostility expressed towards women and femininity."

Martino (1999: 244) reports on similar male attributes in his study of 15- to 16-year-old West Australian males:

"On the basis of this research, it appears that many boys learn to establish their masculinity in opposition to femininity (see Connell, 1994). In other words, they define their masculinity within a set of cultural and social practices which involve a rejection and denigration of what they consider to be feminine attributes or behaviors that often serve as markers of homosexuality in the policing of ascendant forms of masculinity (see Ward, 1995; Butler, 1996; Laskey and Beavis, 1996). This is also reflected in the tendency of many boys to avoid expressing their emotions which appears to be predicated on the basis that showing emotions is considered to be feminine. And as one boys stated in an interview, learning to be masculine involves 'staying away as much as possible from being a female'."

"Foregrounding the primary school as a key cultural arena for the production and reproduction of sexuality and sexual identities, this article goes some way to addressing what are absent from many sociological portrayals of young children and schooling… Discourse of homophobia were expressed vehemently by boys [Grade 6, England] who did not engage in over heterosexual boyfriend/girlfriend relationships and more frequently than by boys who did 'have girlfriends' and who were 'going out'. [Noted are behaviours equating to the] "advocacy of 'queer bashing'. Aside from homophobic narratives communicated in the group interviews, homophobic performances infiltrated classroom and playground interactions. These were directed at boys who got too close to other boys and those boys who failed or chose not to access hegemonic masculine discourses/practices. Differentiating oneself, and subordinating homosexuali