Categorical Dilemmas: Challenges for HIV Prevention among Men Who Have Sex With Men and Transgender Women in Vietnam

Abstract

In Vietnam, HIV continues disproportionately to affect men who have sex with men and transgender women, and the increase in HIV prevalence in these populations may be related to a lack of tailoring of current prevention approaches, which often fail to address social diversity within these populations. To effectively respond to HIV in Vietnam, it is imperative to identify sub-populations within the broad category of ‘men who have sex with men’ (MSM), a term which in Vietnam as in many other sites frequently subsumes transgender women. In this paper, we document the different categories used to describe people who engage in same-sex sexual practices and/or non-normative gender performances drawing on data collected via in-depth interviews and focus groups with a total of 79 participants in Hanoi. We identified over 40 different categories used to describe men who have sex with men and/or transgender women. These categories could be described as behaviourally-based, identity-based, or emic, and each carried different meanings, uses (based on age and geography) and levels of stigma. The categories shine light on the complexity of identities among men who have sex with men and transgender women and have utility for future research and programming to more comprehensively address HIV in Vietnam.

Introduction

HIV among men who have sex with men and transgender women in Vietnam is a significant public health issue that has not been inadequately addressed. Existing research has highlighted high levels of vulnerability to HIV among men who have sex with men, largely due to sexual risk taking behaviours and low rates of HIV testing (Colby, Cao and Doussantousse 2004; Nguyen et al. 2008; Sarraf 2010; Pham et al. 2012; Garcia et al. 2014). Research has also documented elevated risk behaviours among transgender women¹ in Vietnam (Pham et al. 2012; Vu et al. 2016; Clatts et al. 2016). The past decade has shown a dramatic increase in HIV prevalence among men who have sex with men in Vietnam, from 3.9% in 2013 to 7.3% in 2016 (Vietnam Ministry of Health 2014, 2018). An Internet survey conducted with over 2,000 Vietnamese men who have sex with men in 2011 found that 45% reported condomless anal intercourse during their last sexual encounter and less than one-quarter had ever taken an HIV test (Garcia et al. 2014).

In order to address HIV among men who have sex with men and transgender women in Vietnam, it is critical to identify different populations within the broad category of ‘MSM’ and to examine potential variation in vulnerability to HIV (Blanc 2005; Le et al. 2016). The term MSM began to be used in Vietnam, as in many other countries, as part of the early response to HIV and AIDS in the 1990s, and its use became increasingly widespread as a result of the HIV response over the course of the next two decades. Initially, there was little or no recognition of transgender women as being distinct from men who have sex with men, nor was the ‘trans’ category distinguished from the MSM category in many epidemiological statistics. In the 2000s, as transgender gradually began to be separated out of MSM in global AIDS discourse, it also began to separate out in Vietnam, especially among those working with non-governmental organisations. This shift was reflected in relation to official government HIV and AIDS programmes and in the scale-up of the AIDS response in Vietnam that took place in the late-2000s and early 2010s as a result of investment on the part of donors such as the US President’s Emergency Plan for AIDS Relief (PEPFAR) and the Global Fund for AIDS, Tuberculosis and Malaria (GFATM) (Horton, Rydstrøm and Tonini 2015). That said, the amalgamation of these categories continues to be a problem in HIV research (Baral et al. 2013) and the continued tendency to subsume the category of transgender women into the category of MSM in public health research neglects the unique vulnerabilities affecting this population, which include discrimination in healthcare settings (Do et al. 2018) and the use of non-medically prescribed hormones (Nguyen 2019).

In addition to distinguishing between men who have sex with men and transgender women, it is also important to recognise that a range of non-normative sexual and gender identities and behaviours exist within the category of MSM (Aggleton and Parker 2015; Parker, Aggleton and Perez-Brumer 2016). Researchers have noted the challenges of designing effective interventions for such a heterogeneous group (Perez-Brumer et al. 2016). Moving forward, it is critical to understand the diversity among populations of men who have sex with men and transgender women in Vietnam in order to effectively implement HIV prevention programmes. This will entail identifying sub-groups of MSM at high vulnerability for HIV and poor HIV-related health outcomes (Hirsch et al. 2015; Ngo et al. 2009).

Against this background, this article explores the ways in which different categories describing people who engage in same-sex relations and gender non-conforming practices have been defined and classified in different social and cultural systems that co-exist in contemporary Vietnamese society. We explore the problem that these multiple systems of categorisation create for effective HIV prevention programmes, without assuming that there is a simple way to solve it. Based on our exploration of the social and cultural meanings condensed within these categories, we highlight the kinds of issues that will need to be addressed in order to develop future meaningful programmes for HIV prevention.

Methods

The data from this study come from an ethnographic study of men who have sex with men and transgender women in Hanoi, Vietnam. The aim of the study was to collect information on sexual diversity among men who have sex with men and transgender women; challenges and strengths of being Vietnamese and MSM/transgender, and HIV risk behaviours and intervention needs. A total of 72 men who have sex with men and 7 transgender women participated in either in-depth interviews or focus group discussions between March – May 2015. They were recruited through participant observation, outreach activities and through social media (e.g. public and hidden Facebook groups). Thirty-two individuals participated in in-depth interviews and 47 participated in focus groups. Both interviews and focus groups lasted 1–2 hours and were conducted in a private office by trained interviewers. Participants ranged in age from 18 – 58 years. The average age of participants in focus group discussions and in-depth interviews was 27.66 and 30.16 years respectively (see table 1).

Table 1.

Demographic characteristic of participants

Classification Variable		Frequency	Percentage
In-depth interview participants (N = 32)
Age (Mean = 30.16; Range: 19 – 58)	18 – 24	14	43.75
	25 – 35	11	34.38
	Above 35	7	21.87
Self-identified sexual identity	Gay	12	37.50
	Hidden gay	9	28.13
	Conspicuous shadow/trans	5	15.63
	Seme/centre	2	6.25
	Come-out gay	1	3.12
	Heterosexual man	3	9.37
Place of birth	Hanoi	11	34.38
	Red river delta, excluding Hanoi	10	31.25
	Northeast	3	9.37
	North Central Coast	3	9.37
	Southern	1	3.13
	Missing (3 mentioned coming from a rural area)	4	12.50
Focus group discussion participants (N = 47)
Age (Mean = 27.66; Range: 18 – 48)	18 – 24	19	40.43
	25 – 35	21	44.68
	Above 35	7	14.89

Interviews and focus group discussions covered topics including: (a) experiences as a sexual and/or gender minority in Vietnam; (b) perceptions of gay/queer/transgender communities; (c) challenges and strengths to being a sexual and/or gender minority in Vietnam; and (d) HIV-and sexual health-related needs. All interviews and focus group discussions were conducted in Vietnamese, and digitally audio recorded with participants’ permission.

Interviews were transcribed verbatim in Vietnamese. Qualitative data were entered into Atlas.ti 7.0 to assist with organisation and coding of the data. Based on interview guides and open coding, a codebook was developed and used to analyse the qualitative data. Coding was conducted by a team of three Vietnamese researchers and inter-coder agreement across transcripts was greater than 80%. Data analysis relied on the use of grounded theory (Glaser and Strauss 1967), as well as interpretive approaches to the analysis of cultural categories (Geertz 1973; Parker 2009). The Institutional Review Boards of Columbia University Medical Center (IRB AAAM0853) and Hanoi Medical University (133/HĐĐĐĐHYHN) approved all research activities.

Findings

Our research reveals at least three distinct systems of classification—each with its own set of categories (some of which overlap)—used to describe same-sex sexualities and non-normative genders in ways that are relevant for HIV prevention programmes and practices. These were (1) a set of epidemiological and/or behavioural categories introduced for the purposes of responding to HIV and AIDS; (2) a set of identity-based categories that emerged primarily from evolving social movements; and (3) a set of culture-based labels (that we describe as emic or indigenous categories) that are used in daily life and popular culture in Vietnam.

Each of these systems has developed historically and each is typically used in different contexts and/or by different actors. They can also be considered to be intersecting and interacting within evolving non-heterosexual subcultures in Vietnam, and globally. In the following section we describe each of these systems of classification in detail, and the relationships that exist between them to highlight the reasons why paying careful attention to both their differences and their intersections will be essential to any attempt to develop meaningful responses to HIV and AIDS in contemporary Vietnam.

Table 2 provides examples of different terms to describe men who have sex with men and transgender women that emerged in our work. In total, over 40 different terms were mentioned by participants. The sexuality-and gender-based meanings of these terms, and the level of stigma attached to terms, varied widely. However, generally speaking, behavioural and identity-based terms appeared to be less stigmatising than many of the emic, culturally-based terms. Additionally, within emic categories and terminologies, the use and knowledge of terms appeared to be heavily influenced by age and geography.

Table 2.

Examples of behavioural, identity-based, and emic categories used to describe Vietnamese men who have sex with men and transgender women

Vietnamese Term	Literal Translation	Meaning	Note
Behavioural Categories
MSM	Men who have sex with men [English]	A man who has sex with other men	Neutral term
Bot/Bottom	[English]	A homosexual man who plays the receptive role during anal or oral intercourse	Less stigmatising than other terms
Top	[English]	A homosexual man who plays the penetrative role during anal or oral intercourse	Less stigmatising than other terms
Centre/Versatile	[English]	A homosexual man who is comfortable to play both of penetrative and receptive roles during anal or oral intercourse	Less stigmatising than other terms
Chuyển giới	Transgender	A person who is born male biologically but identifies as and/or presents themselves as females.	The term is used in some contexts but has not been universally adopted as a trans women self-identify
Identity-based Categories
Gay/Gay kín		A man who identifies as romantically and/or sexually attracted to other men	Less stigmatising than other terms
Đồng tính	[English]	A man who identifies as romantically and/or sexually attracted to other men	Less stigmatising than other terms
Bi/lưỡng tính	[English]	A man who is sexually attracted to males and females
LGBT	[English]	Lesbian, gay, bisexual and transgender
Trans/chuyển giới	[English]	A person who is born male biologically but identifies as and/or presents themselves as females.	The term is used in some contexts but has not been universally adopted as a trans women self-identify
Emic Categories
Bóng	Shadow, reflection	A man who has feminine characteristics and behaviours	Abusive, insulting terms; more commonly used in the south
Bóng kín	Hidden/secret shadow	A homosexual man who has masculine appearance, but is discreet about his sexual identity	Used in the south and the north
Bóng cứng	Tough shadow	A homosexual man who has masculine appearance but is discreet about his sexual identity. If he does not disclose, it is impossible to guess or know about his sexual identity	Used in the south and the north
Bóng lộ	Conspicuous shadow	A man who has feminine appearance and behaviours, may be transgender. Also used to describe a homosexual man who has come out and open about his sexual identity.	Used in the south and the north. The latter meaning of the category is more common among under-35-year-old MSM.
Seme/Uke/Seke	[Japanese]	A man who is comfortable to play a man’s (seme), female’s (ume), or both (seke) roles in a homosexual relationship	From Japanese literary works about homosexual love, such as gay manga and anime
Bóng mén	Loving/tender shadow	A teenager ‘bóng’	More common in the south than in the north
Bóng chúa	Queen shadow	A middle-aged ‘bóng’ who has come out and has a wide gay/MSM social network and an established reputation in the community	More common in the south
Đồng tính	Of a same sex	Gay, homosexual	Less stigmatising than other terms
Pê đê	Pede [French]	Gay, homosexual man	Abusive, insulting terms
Thế giới thứ 3/ Giới tính thứ 3/3D	The 3rd world; the 3rd sex; 3D	Someone who could not be classified as either a man or a woman; a homosexual man	The term is usually used in social media to describe LGBT people, including homosexual men. It is used by LGBT to describe themselves to non-gay people. Non-gay people also use it to call a person whom they are not sure about his/her gender
Thân sâu mình bướm/Thân sâu hồn bướm	A soul of a butterfly in a worm body	A man who has a woman’s soul, and wants to be a woman	Mostly mentioned by participants over 35 years old
Chuyển giới/trans	Transgender	A transgender female who uses female hormones or has undergone sex reassignment surgery; or an individual, who has a male body, but feels that he is a female.	The use of the term varies. A number of participants think that transgender describes a person who has undergone sex reassignment only, while others understand the term to include a person whose gender identity differs to the one, they were assigned at birth (as it is in the original English term).
Nửa ông nửa bà/ Nửa trống nửa mái/Nửa cái nửa đực	Half man and half woman	Someone who could not be classified as either a man or a woman	Abusive, insulting terms
Đồng cô/Đồng cô bóng cậu	Woman-like man	Someone who could not be classified as either a man or a woman, usually used to refer to a man who is feminine or a transgender woman	Commonly used by older adults in the North. Refers to a spirit medium, wearing a woman’s outfit in ritual ceremonies of the mother goddess religion in Vietnam. It has been expanded to imply a gay man or transgender woman.

Behavioural and epidemiological categories

As a result of sustained public health interventions in the realm of HIV, a range of primarily behavioural categories have come to be used in Vietnam, both officially and also in popular culture, to describe sexual and gender diversity. Participants in this study utilised a number of different behavioural categories to label their sexual and gender identities and experiences. These terms seemed to reflect predominantly western notions and epidemiological understandings of homosexuality. For example, Hung (41 years old) who described himself as a ‘hidden’ (i.e. not out) gay man, suggested that the term MSM was neutral and best for describing men. He noted, ‘the term is the most appropriate one—it’s quite western.’

Other participants—notably those younger in age—spoke more in terms of preferred sexual roles and practices, which in some ways conveyed implicit messages about gender roles. For example, ‘tops’ were described as those who ‘play the husband role’ and prefer the insertive position. ‘Bots’ or bottoms were said to ‘play the wife role’ and prefer the receptive position, while ‘centre’ men were ‘those who can be flexible to play both roles of husband and wife’. According to study participants, these terms were seen as having emerged somewhat recently, and they were only mentioned by young participants (those less than 25). For example, Huong (22 years old), a trans/bot-identified participant noted, ‘[Top and bottom] have been used quite some time, since 2000. Because, in sexual intercourse, it always has a receptive one and a penetrative one’. Quoc (23 years old) described himself as ‘centre’ and suggested that terms like top, bot, and centre provided flexibility in how MSM can describe themselves:

… it means a person can be either top or bot… that you can play both roles, either husband or wife. If you meet someone older than you, so you will be a bot, and your partner is a top in the relationship. If your partner is younger than you and wants to be a top, you will play as a bot, and vice versa.

Another participant named Binh (24 years old), who described himself as seke, a term that derives from Japanese’s Yaoi culture and is a contraction between seme (who plays man’s role in a homosexual relationship) and uke (who plays woman’s role), implied that this label allowed himself to be gay but to maintain his masculinity. In explaining his chosen identity label, he said, ‘I still want to be a man. I am gay, but I want to be a man’. Thus, these participants statements suggest that though top and bottom are behavioural categories, they are quite different from the term MSM in the sense that they reflect cultural particularities in gender and age-based relations. Notably, participants indicated that those who accept a top label were usually older and assume the husband/male role, while those who accept the bottom label tended to be younger and to assume the wife/female role.

The term chuyển giới (transgender) was used by study participants to describe men who perceived to perform and behave as women. As Lan, a 26-year-old participant, said that chuyển giới are men who have undergone sex reassignment surgery to become a woman; or men, who have not yet done the surgery, loves to cross-dress and act as a woman’. Another participant also confirmed that chuyển giới are men who love other men, but who think that they are women. It is a female spirit in a male body. And they use medical intervention to become a real woman as their wish’ (Mai, 27 years old). Notably, several participants mentioned that being (or being identified as) chuyển giới exposes people to more severe stigma and discrimination than sexuality-related categories.

Identity-based categories

Social change and globalisation have brought about an explosion of new categories used to describe the same-sex and non-normative sexuality. Notably, the acronym LGBT—standing for lesbian, gay, bisexual, and transgender—has become an international term for describing sexual and gender minorities globally. In Vietnam, this new language arrived slightly later than many of the terms associated with HIV (such as MSM) —primarily in the mid-to late-2000s, and even more forcefully in the 2010s.

Many participants in our research described themselves in terms of identity-based categories. The labels ‘gay’, ‘homosexual’, ‘bisexual’, ‘LGBT’ (or ‘LGBTQ’) and ‘transgender’ were each employed by participants to imply a social connection to a larger community of individuals with non-normative sexualities and genders. These labels also appeared to convey less stigma than many of the more emic categories discussed below. Additionally, the use of identity-based labels allowed some participants to avoid labelling themselves in constricting terms. For example, Cuong (33 years old) who described himself as homosexual stated: ‘For me, it [categorising myself as a member of a sub-group] does not matter. Whatever it is. It’s not important to me.’ Other participants considered using identity-based categories to label themselves a way to maintain both homosexual and masculine identities. Toan (24 years old), a hidden-gay-identified participant, described gay men as:

Normal [men]. They still have male appearance. There is no difference between their biological sex and their performance. They are still male, perform as male, but they are attracted to other males.

The theme of identity-based labels as a ‘normal’ category was pervasive. Vu (39 years old) suggested that his gay identity ‘normalised’ his life and behaviours, stating:

I’m a gay, and I’m homosexual, but my life is quite peaceful. I do not try to show that I’m gay or do something to let people know [that he is gay]. I just live, normal, just have a normal life, wearing normal outfit, and do not wear any makeup. I just live as what I am.

Likewise, when asked about his self-identification, Quang (35 years old) noted: ‘I never ever think about it. I just live in my way, normal and comfortable… I have never ever had to categorise myself to anyone’. Like many others, this participant identified as bóng kín (hidden gay) and did not disclose his sexual orientation to family or friends. However, he said that he planned on coming out to his father so that he would no longer be pressured to get married.

Other gay and hidden gay-identified men expressed the desire for a normal, non-categorised (or stigmatised) identity, but felt at least some need to hide their identity from family members, who were sometimes said to understand homosexuality as a physical disorder. As Binh (24 years old), a gay-identified participant who had come out to his family reported, ‘their [my parents’] responses were really negative—because they think that it’s a disease that needs treatment’. Nam (32 years old), another participant who identified as hidden gay, expressed similar sentiments—he had not come out to family members but knew that his sister thought that homosexuality was ‘an infectious disease’.

The term ‘transgender’ bridged identity-and behavioural-based systems of classification. While it was predominantly understood in terms of an identity category that connected the bearer to a broader community, it was also used to describe gender performance and behaviours tied to being a woman. Although some participants employed narrow definitions to describe transgender [e.g., one trans-identified participant named Duc (25 years old) said, ‘transgenders are those people who are using hormones or [have] undergone surgery to become a woman, biologically’], it seemed that in practice the ‘transgender’ term seemed to encompass multiple groups of individuals. For example Toan (24 years old), who identified as gay, also alluded to a range of identities and behaviours in the transgender label, noting:

[Transgender refers to] people who were born as male but think that they are female…they might have not undergone sexual reassignment surgery, but they think that they are female, act and behave as a woman. So, they are still transgender. Because many trans people cannot afford to do the surgery, they cross-dress.

Being identified as (or understood to be) transgender opened people up to higher levels of stigma compared to individuals using other identity-based categories to describe themselves. In this way, the western label of ‘transgender’ did not appear to shield those using it from stigma. One 22 year-old transgender participant named Huong spoke of the numbing effect that came with the experience of ongoing stigma:

My friends say that we have been teased since childhood, so we get used to it… People are teasing us a lot, call us by these names so frequently that we get used to it, and do not feel hurt anymore. And now, even us, we are using these terms to make fun of ourselves, to call each other like: oh, you are pê đê [pede], you are đồng cô [womanlike], xăng pha nhớt [gasoline mixed with motor oil], or hai phai [hi fi]. It’s all ok for us. But, it’s only okay for us, who are really out, cross-dressing to perform on catwalks for a long time. For other gays, they will feel awkward [or] uncomfortable if someone calls them [something] like that.

As noted by this same participant, individuals who identified as transgender were ‘really out’ and forced to contend with stigma from unaccepting others, more so than other participants who would be identified using other behavioural or identity-based categories. In addition, the participant’s quote suggests that, in being stigmatised, some transgender individuals may co-opt and reinterpret the language used to marginalise them.

Emic categories

The behavioural and identity-based categories used to define HIV interventions and programmes do not necessarily fit what might be described as an indigenous or popular culture categories that label and describe sexual and gender minorities. Data collection uncovered a range of labels that fit under the emic category, and highlight the complexities, fluidity and temporality of sexual and gender identities among MSM and transgender women in Vietnam. Frequently-used emic terms that describe MSM are bóng kín [hidden shadow], bóng lộ’[conspicuous shadow] seme, and pê đê, among many others. These categories encompassed broad notions about sexuality, gender embodiment and performance, age and geography.

The categories bóng kín and bóng lộ were employed as a way to convey information about sexual preference, but also gender performance. Participants suggested bóng kín were described as ‘so discreet that you will never know’, ‘really manly’ and ‘do not dare to show it’. Binh who was 23 years old defined bóng kín as:

Those who do not want other people know about their sexual identity. People in Hanoi say that bóng kín is a homosexual man who lives and behaves as a normal man because he does not want people figure out his homosexual identity. Their gestures, their styles are quite manly.

While masculinity and discretion were key features of bóng kín, there appeared to be some understanding that men in this grouping were not necessarily excluded from gay or MSM-identified spaces. Duc who was 25 and described himself as ‘belong[ing] to the group of MSM that have come out to family’ noted that bóng kín:

… still go to gay bars quite frequently, but they are quite hidden and secretive in other places. They have nothing different from other normal men, so you cannot know whether they have been infected [with homosexuality].

Like bóng kín, the term bóng lộ carries sexuality-based and gendered meanings. But, in contrast, members of the bóng lộ group was considered feminine and more socially ‘out’, and sometimes the bóng lộ term was used to identify transgender women. Nam (32 years old) noted that ‘bóng lộ is a person who cross-dresses, [who is] out in public spaces’ and that bóng lộ ‘are those people who cross-dress, wear makeup, have feminine appearance and voices. In short, they behave as a woman.’

While the bóng lộ category seemed to focus on transgender individuals, there was great diversity within this label. Several participants suggested that bóng lộ included two groups of people. Quoc who was 23 years-old and identified as ‘centre’ (in English) suggested that bóng lộ includes two sub-groups: ‘The first sub-group include men who have a masculine appearance and wear men’s clothing but have feminine gestures…[the] other sub-group includes men who are cross-dressing and wearing makeup as woman.’ Another participant, Duc (25 years old) suggested that the bóng lộ category included two groups, but gave a slightly different description. He said the first group were those who ‘express it [their sexuality] clearly by their outfit, wearing makeup as woman, or having not very manly gestures’. He continued:

The second group includes those who have come out. They come out to family, friends, and to the whole society, so they do not have to worry anymore.

Thus, there was significant diversity in understanding of the term bóng lộ, with participants suggesting that the category encompasses a broad range of gender and sexual presentations, from men with a feminine gender performance to transgender women who lived their lives socially and physically as women. Even within the transgender sub-group, participants noted that the bóng lộ term includes men who wear makeup and/or women’s clothing as well as transgender women who have taken hormones and/or engaged in gender reassignment surgery.

Emic categories went far beyond bóng kín and bóng lộ. Other culturally-based identity terms frequently employed by participants included the Japanese-originating terms seme, uke, and seke, which implied being a man who is comfortable playing the masculine role, feminine role, or both roles, respectively. These terms are closely associated with Japanese manga (comics) – and specifically with Yaoi, a sub-genre that focuses on same-sexual interactions between boys and young men, that began in Japan in the 1970s (McLelland 2000), and which has circulated globally since then via both print and online media (Wood 2006; Berry et al. 2003).

Though many of the emic categories that emerged in this study were used at times in stigmatising ways, certain terms seemed to be more explicitly demeaning to men who have sex with men and transgender women. The most frequently used derogatory label was pê đê’ (derived from the French word pédéraste, meaning pederast). This label was described as having a ‘scornful meaning’, being used ‘with bad intention’, and a way that ‘people stigmatise [men who have sex with men]’. Other participants pointed out the complexity of the use of this (and other) terms used to describe men who have sex with men. For example, Vu (39 years old), a gay-identified participant who managed a homoerotic massage parlour noted:

In [the past] 2–3 years, when clients text me to ask about my staff, they usually ask whether the staff member is a straight man, gay, or shadow. Recently, no one uses the term pê đê. Before, people did use the term pê đê… I think that people tend to use more polite terms because the education level is getting higher. [‘Gay’ is used] only in big cities, in metropolitan areas. In remote or rural areas, people still use the term pê đê.

Other participants confirmed the existence of educational and regional differences in the use and meaning of the term pê đê and others based on geography and class, suggesting that even terms with a high level of stigma tied to them still has a social context and meaning that must be understood.

Emic Categories as a Window into Health Vulnerabilities and Risks

Emic categories were clearly much more pervasive in the everyday lexicon of Vietnamese men who have sex with men and transgender women, as well as to their families, friends, and communities. Emic terms also tended to be convey more information; in contrast to the term MSM, which usually referred specifically to behaviour, bóng kín and bóng lộ are terms that speak to gender presentation and sexual practice. They also appeared to blend together notions of gender, class, and rural/urban geography, and were not neatly homogenous.

Paying attention to the social and cultural meanings condensed within these terminologies can help to highlight the kinds of issues that will need to be addressed in order to develop meaningful programmes for HIV prevention, treatment and care. These issues pertain not simply to sexual behaviours and identities, but to social inequalities including gender power hierarchies, class differences and the rural/urban continuum. As suggested by existing research, it is plausible that men who have sex with men and transgender women face different kinds of health vulnerabilities according to various dimensions of social inequality, including gender performance (Garcia et al. 2016; Philbin et al. 2018) -where masculine gender presentation has been associated with fewer health seeking behaviours (Garcia et al. 2016), and also class and age-based differences (Philbin et al. 2018).

Another implication regards stigma. As discussed, all of these terms are stigmatised to different extents, with emic terms generally being more stigmatising than either the behavioural or identity-based categories, albeit depending on social context. One difference between men who have sex with men and transgender women in this regard seems to be the extent to which western identity-based categories seem to normalise behaviours and identities. While the identity-based categories gay or bisexual seemed to protect the bearer from stigma, the western label ‘transgender’ did not seem to have this protective effect, perhaps reflecting different levels of social discrimination towards non-normative sexualities and non-normative genders, respectively. It is important to emphasise, however, that although stigma is related to labelling, stigma and labelling are not the same thing. As a social process, stigma extends beyond social labelling and categories to encompass power relations that operate along axes of social inequality (Parker and Aggleton 2003). Therefore, experiences of stigma among men who have sex with men and transgender women may be contingent upon various social factors other than the kinds of labels and categories they apply to themselves, or that others apply to them. Combatting stigma will therefore entail more than simply changing the language and terms that are used, but should also address social inequalities. There is a deep history behind the collective action through which gender normative homosexuality has shed its stigma in metropoles of the global north, at the expense sometimes of rejecting queerness, and it is worth considering that the diverse ways in which these categories are or are not stigmatised—and not just the categories themselves – needs to be examined in relation to global cultural flows.

Another important finding that emerges from a careful reading of these categories—and from paying attention to the ways in which they are deployed in the lived experience of diverse men who have sex with men and transgender women—is the extent to which they are profoundly situational. People deploy different categories in different situations, both when they about and classify others, and when they speak about and classify themselves. At any one moment, and in one particular situation, bóng kín and/or bóng lộ might be the terms of choice, while in a different situation and moment, the terms gay, homosexual, transgender, or any number of options within the broad term ‘LGBTQ’ may be employed by (or ascribed upon) Vietnamese men who have sex with men and transgender women. So rather than being fixed and static, categories are often situationally dependent. This underlines the urgency of focusing on the situations that may create risk and enhance vulnerability for Vietnamese men who have sex with men and transgender women—and the ways in which categorical choices that they make in their lives may have implications for the ways in which they manage or navigate situational risk and vulnerability. It also implies that the classification systems used to identify risk groups and vulnerable subpopulations within communities are useful in describing static moments in a life course, and not always consistent with the lived experiences of sexual and gender minorities in Vietnam.

Conclusion

Thinking about gender and sexuality categories as providing windows into the social realities of peoples’ lived experiences and seeking to identify the connections between these categories and the ways in which they may shape vulnerability and risk, offers insight into issues that must be addressed in order to develop programmes for HIV prevention, treatment and care for sexual and gender minorities in Vietnam.

Our research uncovered a variety of terms used to describe men who have sex with men and transgender women in Vietnam, and our analysis of these terms points to the associated meanings, stigma and potential vulnerability of those whose lives navigate across this terrain of meaning. The categories we uncovered here are not likely to be directly useful in epidemiological analyses, and we are not suggesting that epidemiological categories should somehow be based on the wide range of identity-based of emic categories observed here. That said, HIV prevention and programmatic efforts could be improved by moving beyond birth-or sex-based designations of sexuality or gender, and beyond behaviourally-descriptive categories, to develop more comprehensive and inclusive taxonomy of terms to describe Vietnamese MSM and transgender women, and the sub-groups that exist within these broad terms. Notably, the point of incorporating these local systems of classification is not simply to ensure that programmatic targeting resonates with how local populations actually view and understand themselves, or that measurement more accurately captures people; more importantly, it is to ensure that HIV interventions are better able to address the kinds of social inequalities experienced by different people.

The importance of a range of structural inequalities—related to gender, class, rural/urban experience and notions of modernity in relation to lifestyles and values—also emerge as key issues that must be addressed in the design of future intervention programmes. A key challenge that these complex systems of classification thus pose is how interventions and prevention programmes might address different axes of inequality. One way to do this might be to tailor HIV prevention strategies to different ideologies of gender and masculinity, or to different concerns regarding discretion, or to different desires regarding social mobilisation, or to the different needs of rural and urban, younger and older populations. Careful incorporation of the indigenous, popular categories and classifications presented here could assist HIV prevention better serve sexual and gender minority populations in Vietnam. HIV programmes in Vietnam that fail to consider diversity and inequality as key factors in the lives of men who have sex with men and transgender women will likely to be unsuccessful in reducing risk and could actually enhance vulnerability. This is a point worth emphasising, given that the HIV epidemiological categories in current use internationally (i.e. MSM and, increasingly, transgender women) have shown themselves to be largely inadequate in terms of guiding effective prevention programming and intervention (Colby, Cao and Doussantousse 2004; Nguyen et al. 2008; Sarraf 2010; Pham et al. 2012; Garcia et al. 2014).

Equally important, the situational variability that can be found in the use of categorical alternatives—alternatives that co-existed in the lives of research participants—calls attention to the importance of addressing contextual and situational factors that structure risk and vulnerability in social life. Those who identify themselves as “MSM” while attending an HIV-related meeting may in fact feel more comfortable with bóng kín’/’bóng lộ categories while hanging out in a small circle of close friends. Then they may feel more comfortable to be identified as gay or trans while having sex with another man who prefers to use those identify-based terms. These shifts in using categories to identify themselves may not reflect their rationalized choice but rather their relational one. Such relational shifts may or may not involve changes in their risk practices. It is, therefore, important to acknowledge the contingent aspect of categories while probing further into whether and how context, situation and relation may exacerbate HIV risk.

In sum, careful attention to indigenous, popular categories and classifications provides important insights into the landscape of sexual and gender diversity. Moreover, greater understanding of this diversity helps us to better develop and implement programmes that can more effectively serve sexual and gender minority populations in Vietnam. More generally, the work undertaken here exemplifies the importance of attending to diverse and fluid sexual cultures, as surfaced through the labels that people use to map their own worlds, in all the places where sexual and gender minorities face substantial health disparities—in other words, everywhere.