Gay Language Indonesia: Registering

Many homosexual men in Indonesia speak what they call bahasa gay ‘ gay language’, a linguistic phenomenon based upon bahasa Indonesia (Indonesian), Indonesia’s national language. Bahasa gay involves derivational processes including unique sufﬁxes and word substitutions, and a pragmatics oriented around community rather than secrecy. Although mainstream knowledge of gay men’s existence is limited, bahasa gay is increasingly being appropriated by Indonesian popular culture. By examining bahasa gay in terms of state power and register, the article asks how this form of speaking might contribute to better understanding how gay subjectivity is bound up with conceptions of national belonging. Gay Indonesians might seem to epitomize difference; they seem to lie radically outside the norms of Indonesian societies. Within gay communities and in popular culture, however, bahasa gay appears as a register of belonging, not one of hierarchy or distance. [Indonesia, gay, nation, register, belonging] decentered at its centre of of social differentiation, a that on modes and zones of contact between dominant and dominated groups ... that focused on how such speakers constitute each relationally and in difference, how they enact differences in

. Bahasa gay is an important piece of evidence suggesting that these transformations are also shaped by national discourse.
Bahasa gay, terms from which I mark with bold italic, is also known by gay men and other Indonesians as bahasa banci, a closely related language variety. Banci is a nationwide (and somewhat derogatory) term for male-to-female transvestites; two well-known bahasa gay/banci variants of the term are binan and béncong (thus this language is also called bahasa binan or bahasa béncong). In contemporary Indonesia, male-to-female transvestites prefer the term waria (an amalgam of wanita 'woman' and pria 'man'), and I use waria to refer to these persons for the remainder of this article. I do this not only out of respect but also for analytical precision, because banci can also be a joking term between gay men or a general term of opprobrium, somewhat like faggot in English. Bahasa banci therefore also has the connotation 'waria and gay language' or even 'queer language' or 'faggot language'. It appears that bahasa gay originated as a variant of bahasa waria. Although I have encountered both gay men and waria who insist that their languages differ, the two varieties retain many similarities, are spoken in overlapping everyday social worlds, and borrow from each other, to the extent that they are sometimes treated as the same entity. 2 I, too, view bahasa gay and bahasa waria as essentially the same thing despite minor differences. Thus, although this article seeks to highlight the particular features and implications of bahasa gay as spoken by gay men, my conclusions regarding the impact of the nation-state on gay language may be relevant to waria language as well, since waria are also found nationally (and are not seen as limited to any particular set of ethnicities) and concern themselves with their place in national society (Boellstorff 2004a).

The Idea of an Indonesian National Culture
Indonesia is the world's fourth most populous nation, with over 220 million citizens spread out over about 6,000 inhabited islands and speaking, by some counts, nearly 700 local languages alongside Indonesian. The very thought that there could be a robust and deeply felt national culture in this archipelago might seem incredible when, since the fall of the authoritarian leader Soeharto and his New Order in 1998, religious and ethnic conflict has greatly increased in many parts of the nation, and East Timor, a former Portuguese colony invaded and annexed by Indonesia in the 1970s, has become independent. Yet scholarship on Indonesia (e.g., Brenner 1998;Errington 1998;Spyer 2000) continues to find that the nation is an important, though certainly not exclusive, social arena and that "ethnolocality"-the assumption that identity is always founded first and foremost in an isomorphism between ethnicity and locality-does not tell the whole story (Boellstorff 2002).
In her critique of "linguistic utopias," Mary Louise Pratt (1987) notes that Benedict Anderson's notion of the "imagined community" can be used to characterize the idealized speech communities that, in her view, undergird sociolinguistic inquiry. Pratt uses utopian to mean that nations and language communities "are imagined as islands, as discrete and sovereign social entities" (1987:50). This is a continental understanding of islands as isolated from one another. Indonesia, however, is a nation where sovereignty is linked not to discreteness but to a network of power, enshrined in the state's "archipelago concept," a more institutionalized analog of the melting-pot metaphor of nationhood in the United States. Originating in debates over maritime law, this concept subordinates local island culture to a national identity conceived in terms of an archipelago. By 1973 it was officially declared "to be the concept that forms the basis of Indonesia's national development" (Kusumaatmadja 1982:12). Bahasa gay is also shaped by this national discourse. Throughout my fieldwork on several Indonesian islands (primarily Java, Bali, and Sulawesi), I have looked for regional or local distinctions in bahasa gay. To date I have found only minor and temporary variations. Given the well-documented and extensive variation in local cultures across the Indonesian archipelago, such similarity is particularly striking. This article provides linguistic evidence for the hypothesis that gay subjectivity is bound up with a fractured but real national culture.

Registering Belonging
My analysis does not hinge on examples of gay men using bahasa gay to speak about their sense of belonging to a national community. People cannot and do not always comment explicitly on the cultural beliefs they hold; this is one strength of ethnography over methodologies based on elicitation. To say that language is shaped by x is not the same thing as saying that x is the topic of conversation; gender norms can shape language use even when gender is not under discussion, and the use of ethnic dialects is not limited to talk about ethnicity. After all, "language ideology . . . is only partly captured in what people say about language. People's talk about language is likely to be less nuanced than their practical but tacit understandings that are embedded in how they actually use language" (Keane 1997:98).
Nonetheless, gay men do occasionally comment on their relationship to national discourse. During my 1997 fieldwork in Surabaya, one of the most familiar faces at Texas, the bahasa gay term for a popular meeting place for gay men, was Andri, who was known as a community leader. 3 One night Andri was in the midst of a group of about 20 gay men in the "center" of Texas-a space under one of the few working streetlights on the roadside, where an old woman had set up a small food stall to sell snacks. Andri was engaging everyone in hilarious conversation when he suddenly turned his attention to a space about ten yards down the road. Here a large tree provided shelter from the feeble street lamp, allowing one to be "in Texas" while in darkness. This part of Texas was popular with tertutup 'closed' men who entered the gay community only to find sex partners, and with teenage men not yet comfortable with the ways of the gay world, including gay language. On this night Andri had noticed in the darkness a young man sitting on a motorcycle, whom no one knew. The young man, whom I will call Ali, sat in shy silence, watching the scene with apparent fascination and discomfort.
Never one to let an opportunity for drama slip by, Andri took Ali's shyness as a challenge. Dropping whatever line of commentary had been occupying him to that point, Andri walked right up to Ali, who now appeared positively embarrassed as all eyes turned to see what Andri would do. Andri began with a few short sentences: "Aduh, brondong sekali!" 'My, isn't he young!'; "Cucok kamu ya!" 'You are so handsome!' In each utterance one key term was not in standard Indonesian but in bahasa gay (brondong 'young man' and cucok 'handsome'). Ali responded with a fetching but silent smile; Andri concluded that he did not know bahasa gay, was too shy to speak, or both. Clearly, more effort would be required to obtain the desired entertainment effect. Thus Andri suddenly stood up straight and formal like a government official, turned one fist into an impromptu microphone, raised his voice to just below a shout so all Texas could hear, and, taking on the measured tones of a television reporter, said, "Are you new here?" Ali answered haltingly into Andri's "microphone," "Yes, this is my first time at this place." Andri's voice now shifted to that of a cheery game-show host: "Well, you just keep coming back here, okay? Come back here tomorrow and it will be even busier. This place will make you happy. Just bring your National Identity Card [KTP, short for kartu tanda penduduk] and we'll get you a second, hémong ['homo'] National Identity Card. Just bring some forms and two photographs and 15,000 rupiah and we'll get it for you." The audience burst into laughter, which only increased as Andri pointed to the old woman running the food stall and added: "This lésbong ['lesbian'], who is also a pimp ['germo'], will do it for you." The old woman laughed, accustomed to such jesting. One of the gay men watching Andri's performance shouted out, "What will he get with his new gay National Identity Card?" Andri answered, "A lékong ['male']." In this example, Andri, in a performing mood at Texas, explicitly uses bahasa gay to mark a sense of community-and thereby tries to draw Ali into this community.
Andri also represents the gay world as a kind of alternative national space, complete with an air of officialdom. Although such examples are striking, the place of national discourse in gay subjectivity emerges most often through implicit assumptions about selfhood and belonging.
In this article I approach bahasa gay as a register. This concept is notoriously ill defined, even "pretheoretic," in scholarly analysis (Iwasaki and Horie 2000:519; see also Biber andFinegan 1994:4 andHervey 1992:189). Most early definitions defined register as variation according to use (social context) rather than variation according to user, which was often termed dialectal variation (Asher 1994:3509;Bussmann 1996:402;Leckie-Tarry 1995:6). More recent scholarship on register has revealed how "inference choices drawn from stylistic choice reflect back on information about the languageuser" (Hervey 1992:192); the pendulum has swung from use to user such that registers can now be seen as "independent of . . . parameters" of use (Paolillo 2000:217).
The most productive analysis of this oscillation between use and user is Asif Agha's claim that "diversity of metalinguistic opinion on the part of linguists is motivated, to an extent, by differences of metapragmatic opinion on the part of language users" (Agha 1998:154). Agha's point is that the relationship between social identities and social context-between user and use-is "leaky"; ways of speaking are "stereotypes" about kinds of people and vice versa. In other words, the issue is not to resolve whether registers are either about kinds of speakers or about contexts of speaking; registers linguistically perform and sustain both this conceptual dichotomy and its inevitable "leakage" in actual practice. This formulation recalls not only work in the anthropology of sexuality that exposes the contradictions of the identity/behavior binarism (Elliston 1995), but also work on language ideology that reveals how beliefs about language reflect beliefs about society (Irvine and Gal 2000;Kroskrity 2000;Silverstein 1979Silverstein , 1998. My analysis draws upon these literatures to suggest that bahasa gay indicates how the lifeworlds of gay men are "leaking" into Indonesian national culture even as most Indonesians remain at best ignorant of, and at worst openly hostile to, their existence. It is the productive tension between context and subjectivity that, as I discuss later, makes bahasa gay so amenable to appropriation by Indonesian popular culture. Bahasa gay suggests that as gay men become better known in Indonesia, their social stereotyping (Agha 1998:168;Hervey 1992:195) is increasingly as members of national culture, not "local" or "traditional" cultures.
Most theories of how power systems shape subjectivities frame this influence in terms of either interpellation through a dominant ideology (e.g., Althusser 1971), or a "reverse discourse" or negative interpellation that is at odds with, yet constituted through, a dominant ideology (e.g., Foucault's 1978 analysis of the European and American homosexual). Bahasa gay raises the possibility that systems of power can instead create structuring conditions for subject positions that they neither call into being nor repress. In such cases, the subject position in question would not be in alignment with or antagonistic toward the ideology to which it would nonetheless be beholden for cultural coherence. In Indonesia, for example, the imbrication of the gay subject position with national discourse can persist regardless of particular gay Indonesians' views on the nation. Bahasa gay may therefore help us see how a system of power can result in subject positions that speak neither with nor against that system, yet articulate their unexpected logics in terms of that system's grammar-literally and figuratively.

The Setting
Many gay men emphasize that they have a way of speaking in the gay world (dunia gay) that differs from speech in typical Indonesian society-what they (and Indonesians more generally) often term the normal world (dunia normal). Then as now, the gay world refers not to a fixed topography but to an imagined geography where gay men can be terbuka 'open'. It comes into being any time gay men gather at the slightest remove from the normal world. This imagined gay world is not contiguous, but composed of an archipelago of physical locales ranging from parks and salons to shopping malls and individual apartments; it also includes intangible romantic, sexual, and friendly networks of affiliation. Physical locales become part of the gay world only when occupied by gay men; during the daytime, Texas is just a small road behind a bus terminal. Some of these places are given names, the motivation for which tends to be idiosyncratic. In Surabaya, where places in the gay world are most commonly given names originating outside Indonesia, such names are learned from mass media (and associated with the West) but often have an additional local motivation. (Texas sounds like a contraction of Indonesian terminal bis 'bus terminal'; Kalifor (from California), which is located next to a river, begins with kali 'river'.) In Yogyakarta the town square is known by many as L.A., which recalls the Indonesian word for town square, alun-alun, while the town square in Solo's center is often termed Manhattan. The town square in Makassar is simply called kampus ('campus'), reflecting its status as a place where men gather away from home, and the term kampus is used in many other cities as well.
In contrast to these cosmopolitan and collegiate images, most gay men are lowerclass and the gay world is-contrary to stereotype-not exclusively urban, although the best-known, largest, and most durable nodes of the gay world are found in cities. It is for this reason that I have focused my research on three urban or semi-urban areas: Surabaya in East Java, Makassar (formerly Ujung Pandang) in South Sulawesi, and the Denpasar-Kuta region in southern Bali. I chose these sites because they were relatively close to one another, yet contrasted in ethnic and religious makeup, degree of contact with non-Indonesians, and position in the Indonesian nation-state. I have also visited eight other urban or rural areas of Indonesia with gay communities, including sites in Kalimantan, Central Java, and the national capital of Jakarta.
Gay men not only informed me of the existence of bahasa gay but also eagerly taught it to me. I likewise observed such men teaching bahasa gay to other Indonesian men who were new to the gay world (such teaching was almost always limited to lexicon). Gay Indonesian men knew me as a self-identified gay U.S. citizen and enjoyed showing me what they saw as an interesting and important part of their world, which they are often happy to show to normal persons, too. These men believed gay men and gay men to be similar in some respects, yet saw the Indonesian gay world as distinct. Bahasa gay has arisen in a context rich in secret languages and specialized argots, the best known of which is prokem, an urban street language based on Indonesian (Chambert-Loir 1984;Dreyfuss 1983;Rahardja and Chambert-Loir 1990;Saleh 1988). Prokem appears to have originated in the criminal underworld of Medan and Jakarta in the early 1960s, spreading to street youth by the late 1960s and to university-educated youth by the mid-1970s (Chambert-Loir 1984. Bahasa gay probably began in the 1970s, and a few items of bahasa gay (such as se' 'homosexual man') probably date back to the 1960s (Oetomo 2001) and were part of the earlier male homosexual worlds that are one important historical source shaping the contemporary gay subject position. The earliest record of bahasa gay of which I am aware is a wordlist of 25 items gathered in the mid-1970s in Yogyakarta by Amen Budiman (1979:106-107); many items from this list are still in widespread use.

The Indonesian Foundation
To date, the fundamental condition of bahasa gay's existence is that although some terms transform words from local languages such as Javanese or Balinese, at the overall grammatical level bahasa gay is always based on Indonesian (bahasa Indonesia), the national vernacular. Given the size of Indonesia's population and its considerable linguistic diversity, I had every reason to expect that as I moved from island to island and between different regions within islands, I would encounter differing gay languages based on, say, Makassarese or Balinese. What I discovered was something gay men Table 1 Lexemes found in the bahasa gay of makassar claimed to be from Javanese.

Term (Origin)
Meaning Replaces Indonesian term aluk (unknown) penis kontol mendolo (Javanese) pop-eyed lihat 'to see' kula (Javanese) I, me, my saya longka (probably Javanese langka 'exceedingly big') big besar muttu (unknown) have sex main seks neka (unknown; possibly Javanese eka 'one') he, she, it dia ora (Javanese) no, not tidak sampeyan (Javanese) you kamu sibollo' (unknown) friend sahabat ta'belo (possibly from Javanese tabel 'puffy, swollen') bad jelek tabu (unknown) eat makan already knew: bahasa gay is a self-consciously nationwide way of speaking. Gay men sometimes explicitly comment on the national character of bahasa gay, as in the case of Eddy, an ethnically Bugis gay man in Makassar, who emphasized that there was no such thing as a bahasa gay Bugis 'Bugis gay language' but only a bahasa gay Indonesia. Eddy noted that local variations in bahasa gay vocabulary existed but that these were all part of a national bahasa gay. Of course, lexical items of bahasa gay that are limited to certain areas can be found in many locales; bahasa gay changes quickly, although some terms and phonological patterns persist for decades. Based on my fieldwork, there appears to be a widespread understanding among gay men like Eddy that such variations are just that, local variations on a bahasa gay that is sama 'the same' across Indonesia. This term sama is also the term gay men use to describe their homosexual desires: they term them a "desire for the same." The common belief that bahasa gay is the same across Indonesia is reinforced by the fact that to date, all of the derivational patterns used to produce bahasa gay lexemes originated in one region of Indonesia but became nationally distributed through gay social networks. 4 One variation on this pattern is that terms that are claimed to be Javanese (though often of unclear origin) have become an element of bahasa gay in Makassar on the island of Sulawesi (see Table 1). This is not simply an instance of the broader incursion of Javanese into the national vernacular (Anderson 1990b), because these terms are not used by gay men on Java itself; their use is distinctive to gay Makassar, reflecting a sense of translocal connection.
The significance of bahasa gay's founding in Indonesian is a consequence of the unusual position of Indonesian in the nation-building project. Language played a vital role in the state's enormous effort to build a sense of nationalism among the denizens of the Dutch East Indies. At the time of independence in the late 1940s, many of these groups shared little more than the Dutch colonial encounter, which itself exhibited great regional variation in length and intensity. One core element of most nationalisms is the belief that to be modern and authentic, nations need national vernaculars (Anderson 1983). What would become Indonesia's mother tongue? Dutch clearly would not have sufficed, not only because of its association with the colonizing power but also due to Holland's antipathy toward having its subjects speak the colonial language. At the time of independence, after 350 years of Netherlands rule, less than two percent of Indonesians spoke Dutch (Anderson 1990a(Anderson :138, 1990bGroeneboer 1998:1;Siegel 1997:13). Javanese, spoken by almost 40 percent of "natives," seemed a logical choice, but selecting any local language had the disadvantage of privileging one group. A solution was found in Malay, "the language of certain courts and of villages, though not the language of the largest groups of the archipelago" (Siegel 1997:14), the lingua franca of the Dutch East Indies. The colonial bureaucracy had inadvertently prepared the way by inculcating a conceptual shift "from heteroglossia to polyglossia" (Maier 1993) wherein other languages (now termed bahasa daerah 'regional languages') were assumed to transparently index distinct groups (Javanese Java Javanese people, Balinese Bali Balinese people, Torajan Torajaland Torajan people). In contrast-fortified by prior centuries of trade in which Malay had become distributed not only across the Indies but as far away as the Philippines, Japan, Sri Lanka, and Madagascar (Errington 1998:52)-Malay was construed as a placeless, peopleless tongue.
Malay's ubiquity reflected the assumption that the colony was a unified entity. By the 20th century, however, a reverse discourse had formed in the sense that the prevalence of Malay, albeit an instrument of colonial power, began to fuel a belief that the disparate ethnic groups of the archipelago could be part of a transethnic and translocal imagined community opposing colonialism. Siegel notes that a lingua franca like Malay "by definition operates between peoples of different languages and cultures without belonging to any of them. . . . Thus, [in the Indies] it produced something that was not completely foreign or completely domestic" Siegel 1997:8-9). That "something" was the Indonesian nation itself.
For Malay to become a language of nationhood, a concerted effort was needed to transform a language of colonial domination into one of national unity, transforming the plural society into a nation by constructing a paradox-an authentic lingua franca. William Liddle notes that in the hands of nationalists and later the postcolonial state, Malay, renamed Indonesian, was made "perhaps the most important single ingredient in the shaping of the modern culture" (Errington 1998:4; see also Anderson 1966;Errington 2000:208;Siegel 1997). Since under colonialism Malay/Indonesian had been contrasted with languages seen as native, to this day its "un-nativeness crucially enables and informs its place in the Indonesian national project" (Errington 1998:3). There is a misunderstanding that Indonesian is an invented language; what has been invented is its speech community. This language frequently appears in grammars as "Malay/Indonesian": the slash simultaneously linking and separating Malay from Indonesian marks a shift not in grammar but in the manner of imagining communityfrom a lingua franca of trade and colonialism to the archipelagic key in which the new nation's authenticity would be played. For my interlocutors in Java, Bali, and Sulawesi, mostly born in the 1960s or later and for whom Soeharto's New Order was all they knew until the era reformasi 'era of reform' beginning in 1998, Indonesian is a feature of everyday life (Errington 2000:209). All of my interlocutors speak Indonesian, as does almost 90 percent of the Indonesian populace, approximately 15 percent of whom now speak it as their first language. This percentage is increasing, and the use of Indonesian as a first language is increasingly linked to middle-class identity (Oetomo 1996). 5 It is a language not only of the public sphere, commerce, and the state, but of family intimacies, romance, and emotion; for most of my interlocutors it is learned from the earliest years of life and in some cases is the only language they speak. This spread of Indonesian, which has been the subject of many studies (e.g., Errington 1998;Kuipers 1998), testifies to the success of the New Order state's educational initiatives: Indonesian was "intimately bound up with the New Order's fortunes, as is clear from one of Soeharto's very first unilateral decisions: a 1965 Presidential Instruction which mandated the government-supervised building and staffing of elementary schools throughout the country, particularly in rural areas" (Errington 1998:59).
As a result, "among the New Order's most enduring effects on the Indonesian landscape [was] its success in propagating Indonesian-ness with and through the Indonesian language" (Errington 1998:2). Bahasa gay, founded in Indonesian, partakes of Indonesian's nationalistic tenor; bahasa gay is also a language "that lacks a primordial ethnic community of native speakers" and is "less nonnative than unnative" (Errington 2000:206). I later return to bahasa gay's founding in Indonesian as part of a discussion of ideology and hegemony. One point among others will be that bahasa gay does not act as an anti-language in Halliday's (1976) sense, because the language's fundamental logic is not that of alterity but of creative transformation of a dominant discourse-namely, state discourse-thus remaining within its horizon. Similarly, it does not fit Morgan's definition of a "counterlanguage" as "a conscious attempt . . . to represent an alternative reality through a communication system based on ambiguity, irony, and satire" (1993:423). This is because although bahasa gay is a conscious (and often humorous) language game, it represents not an alternative reality but a queer take on a dominant reality.

Derivation
Although competence in bahasa gay includes intonation, pragmatics, and ideology about bahasa gay itself, what gay men (and those who appropriate bahasa gay) find most salient is lexicon; indeed, scholarly and everyday commentaries on bahasa gay frequently mention only its lexicon. This "lexicon," however, is more than just a collection of words; like prokem, it is a set of patterned derivational processes that together constitute a language game. In its emphasis on derivational processes, bahasa gay resembles not only prokem but gay languages elsewhere in Southeast Asia and beyond, including Filipino swardspeak and British Polari (Baker 2002;Lucas 1997). However, unlike swardspeak, in which one process, ipis talk, is said by some to make the face look "queenie" due to the frequent [i] and [v] sounds that purse the lips (Manalansan 1995:203), the derivational processes of bahasa gay are not considered inherently effeminizing. It is the existence of these productive derivational processes that typifies bahasa gay; true fluency is signaled not just by knowing vocabulary but by knowing the processes and being able to coin neologisms oneself.
There are several ways to create gay terms, most of which involve retaining the first syllable of a standard Indonesian word and then modifying the ending. Thus most terms in bahasa gay form a kind of commentary on standard language and in this sense resist dominant norms, although such resistance is not always overt (see Oetomo 2001:67-71). 6 Particularly since the mid-1990s, the most popular process is syllabic substitution, where a word replaces a standard Indonesian word with which it shares a syllable (typically the first syllable). For instance, tidak 'no, not' is replaced by tinta 'tint' (see Table 2). This process may have begun in the city of Medan in Sumatra (Oetomo 2001:65). In some cases the substituting word comes from a local language, There is usually no semantic link between the substituted word and the original, and this semantic dissonance is part of the humor of bahasa gay, although weak semantic motivations are sometimes discernible. Examples are kelinci 'rabbit' replacing kecil 'small' (but jelita 'lovely' replacing jelek 'bad') and the standard Indonesian phrase ya ampun ('Oh my God!'; ampun literally means 'forgiveness') being replaced with ya amplop (amplop 'envelope'). Sometimes the substituting word comes from a product name, as in bodrex (a cold medicine) replacing bodoh 'stupid', or from a place name, such as Makassar for makan 'to eat'. This use of place names, seen also in the naming of some hang-out places with geographic terms, such as Texas, Kalifor, and Manhattan, appears to reflect a connection between bahasa gay and the idea of the gay world as a distributed network of locales.
Two other derivational processes are related to syllabic substitution. The first is neologism (Table 3), in which the Indonesian term is replaced by a form that shares the same first syllable or sound but does not have a prior meaning of its own. Only a handful of bahasa gay terms originate in this manner; bahasa gay is a language of transformation. The second process is semantic shift, whereby an Indonesian term is given a new meaning (Table 4). Semantic shifting is a feature of non-gay urban language as well, where it is termed plesetan (Chambert-Loir 1984; Oetomo 2001).
Another important derivational process in bahasa gay is suffixation and vowel shift (Table 5), which is usually used to transform a standard Indonesian term but occasionally involves a bahasa gay item, a local language term, or an English loanword. The most common suffixes are -ong and -es; -i is a less productive variant that arose in the 1990s (Oetomo 2001). In the case of -ong suffixation, the vowel of the immediately preceding syllable of the lexeme to be transformed shifts toé. Dédé Oetomo claims that the same process takes place with the -es suffix, but at least in Makassar I have found that a shift to the schwa, rather thané, sometimes accompanies this suffix. Oetomo suggests that suffixation and vowel shift first appeared in Jakarta and areas most directly influenced by the Jakartan dialect of Indonesian and may be derived from prokem, although the suffixes differ; he also notes that a few terms of -ong shifting, namely béncong, from banci 'male transvestite', and népsong, from napsu 'desire', appear to have come into existence before bahasa gay took form (2001:62). In 1984, Henri Chambert-Loir noted that "transvestites are generally known to use words with the  (1984:110). At least some of these terms (e.g., polésong, lékong, gédong) remain current in bahasa gay. As in the case of all other derivational processes for bahasa gay, the most common kinds of transformed words are nouns and adjectives, but verbal lexemes can be transformed as well, as in the case of keménong, or terjédong for terjadi 'to have happened'. Beyond syllabic substitution (including neologism and semantic shift) and suffixation with vowel shift, there are several minor processes for creating bahasa gay vocabulary through affixation and acronyms. Si-prefixing is primarily found in Javanese-speaking areas (Oetomo 1999(Oetomo , 2001 but, like all elements of bahasa gay, spreads elsewhere. In this process the first syllable of a term is prefixed with si-, other syllables are deleted, and a consonant is added at the end of the word if the syllable ends in a vowel. Si is a Javanese and Indonesian particle indicating categories of persons; in bahasa gay it usually retains this meaning. Examples include silan (from Javanese lanang 'man'), sihom (from homo), siban (from banci 'male-tofemale transvestite'), siG for gay, and siL for lesbi (the letter g is pronounced [gei] in Indonesian; l is pronounced [El]). The town square in Yogyakarta, known in bahasa gay as L.A., is sometimes called Si-A.L.; the L and A are switched in this case to create a pun (sial is Indonesian for 'bad luck').
With -in-infixing, the infix -in-is "inserted between the consonant and vowel of every syllable, usually with a shortening of the product so that it becomes two syllables long" (Oetomo 1999:28). Thus banci becomes binancini, which becomes binan. Linak, a bahasa gay term for 'man', is another example of this process (laki > linakini > linak), as is lines for 'lesbian' (lesbi > linesbini > lines). Oetomo (2001) notes that this process appears to have begun in Jakarta and Bandung but has spread across Indonesia (it is known in Makassar, for instance). Infixing is not unusual in Indonesian wordplay: Chambert-Loir considers the -ok-infix to be "the only bahasa prokem element, the uniqueness of which cannot be disputed" (1984:111-112); the infixes -in-and -ark-/arg-appear in Javanese urban languages, and in the case of -ark-/-arg-, Indonesian ones as well (Chambert-Loir 1984:109). 7 An additional process is -se' suffixing, in which all but the first syllable of a word is deleted (plus the first consonant of the following syllable, if the first syllable ends in a vowel) and the remaining syllable is suffixed with -se': homo > hom > homse'; Cina 'Chinese' > Cin > Cinse'.
Yet another means of forming gay language terms involves the reinterpretation of standard Indonesian terms as acronyms: For instance, kopi susu 'coffee with milk' can mean (ko)ntol (p)anjang (i)tu (s)angatk(u) (su)ka 'I really like that big cock' (Boy and Yasiano 1999:41-43). This process is best documented in the city of Bandung in West Java. In prokem such reinterpretations are a long-standing form of wordplay (Chambert-Loir 1984:107), and I have often heard examples in everyday vernacular Indonesian.

Intonation
Though far less emphasized than patterned lexicon, speaking in what is considered to be an effeminate manner is also sometimes asserted by gay men to be indicative of bahasa gay. By 'effeminate' (standard Indonesian kewanitaan or feminin; bahasa gay terms include ngondhek, mégol, kriting 'curly'), these men refer to the high-pitched tone and rising utterance-final intonation that Indonesians associate with images of demure femininity and softness (lembut) (cf. Gaudio 1994 for perceptions of American gays' pitch patterns). For example, one night at Texas, Karno, a gay man, told me he was preparing for a vacation in Bali with the aim of meeting Western men. (One reason Bali is a popular internal tourist destination is that it provides an opportunity to observe Westerners.) Karno asked, "How can you tell which white people are 'like this' (begini) and which ones aren't?" We discussed some Western gay symbols, and then I turned the question around, asking if he could tell which Indonesian men might be gay. Karno and the other men present agreed that one usually could tell, and Karno emphasized speech as an indicator. About half an hour later I was speaking with another man in the group when Karno interrupted me: "That's it! You're talking how gay men talk right now! Instead of saying 'like that' [begitu] normally, you just said it this way," and he imitated my use of rising utterance-final intonation.

Secret Language or Community Language?
As I gained some competency in bahasa gay, I discovered that I needed to learn not only words, but the pragmatics of their use in everyday interaction. To my knowledge, people are never directly instructed in these pragmatics as they are for lexicon. Indeed, one of the two consciously articulated ideologies concerning bahasa gay proves to be an unintentional red herring: if asked, "Why does bahasa gay exist?" many gay Indonesians respond that it is a secret language (bahasa rahasia), what Michael Aceto (1995) calls a "cryptolect." I am far from alone in encountering this response: For instance, Budiman (1979) refers to bahasa gay as bahasa rahasia; and Richard Howard's study of gay men in Jakarta notes that "individuals explained to me that they use gay slang because they could speak freely about their homosexual desires and experiences without worrying that other people could understand what they were saying" (1996:9). He adds, however, that "the use of gay slang also functions to foster a sense of belonging to a community" (1996:9; see also Oetomo 2001:67). Similarly, gay men in the Philippines explicitly justify their gayspeak or swardspeak as "communicating with each other in a way in which the outside . . . world is unable to make sense of it" (Manalansan 1995:202).
I once heard this understanding of bahasa gay articulated by Linda, a masculine lesbi woman who had learned bahasa gay from gay friends. One day Linda decided she needed to visit a mystical shrine on the slopes of a volcano several hours outside Surabaya; she had lost her job and didn't have a girlfriend and was hoping for some insight on both fronts. I accompanied her, and on reaching the mountainside village where the shrine was located, we settled in at the inn of Sunardi, who knew neither that Linda was lesbi nor that I was gay. The next morning Linda and I saw an effeminate man on the neighbor's porch cutting someone's hair. "That's my younger brother," Sunardi said. "He moved to the city a few months ago and works in a salon there." My suspicions that the younger brother might be gay were heightened about half an hour later when Linda made a joke and Sunardi added "péres," derived from pura-pura 'pretend', a bahasa gay particle meaning 'gotcha!'. Sunardi explained that his younger brother had "learned a new language at the salon in town" and had taught some of it to him. On the ride back to Surabaya that afternoon, Linda was incensed. "Sunardi's younger brother isn't very professional (kurang profesional) to tell everyone about bahasa gay. That's for our group alone, so we can talk without other people understanding." I term this widespread view that bahasa gay is a secret language a red herring not because Linda and others are dissembling. However, while "the theme of secrecy is a familiar one in what we might call 'folk anti-linguistics' . . . it is unlikely to be the major cause of [its] existence" (Halliday 1976:572; see Bolton and Hutton 1995;Goyvaerts 1996). Bahasa gay can act as a secret language in some cases, but its actual pragmatics appear to reflect more closely the second consciously articulated ideology about it, which is provided by both gay men and outside commentators: bahasa gay is a "slang" in the sense of a language of association and community (bahasa gaul). My argument is that it is the goal of association that makes a particular utterance a valid "move" in the game of bahasa gay and that what is at issue in this association is a sexual community understood in national, not ethnolocalized, terms.

Against the "Secret Language" Hypothesis
Several pieces of ethnographic evidence indicate that bahasa gay is not a secret language. The first is that not all gay men know it; they are not all privy to the "secret." Gay men who avoid gay places (for instance, interacting only with a small circle of friends) may have little or no knowledge of this form of language. It is neither a necessary nor sufficient condition for gay subjectivity: persons who do not know a word of bahasa gay can identify as gay, and non-gay Indonesians who spend time in gay places (e.g., waria, lesbi women like Linda, female sex workers, pedicab drivers, salon workers) can become proficient in it. Bahasa gay could thus be at best the secret language of a subset of gay men, and it could not serve to keep secrets from precisely those non-gay Indonesians who spend time in and around gay communities.
Second, in contrast to language games such as Pig Latin, whole clauses of bahasa gay are rare. As noted below, when so often only a single word in an utterance is changed to bahasa gay, and that word shares a first syllable with the standard Indonesian term it replaces, the resulting utterance is not very secret. Occasionally every word in an utterance will be in bahasa gay, as in the following examples: (1) Standard Indonesian: Bahasa gay: (2) Standard Indonesian: In Example 1, the standard Indonesian phrase Aku tidak mau 'I don't want' is replaced with Akika tinta mawar ('I [neologism] tint rose'). The effect is roughly like what an English speaker would hear if "I don't want" were replaced with "Eyesore donut wonton," or if "That boy is cute" were replaced with "That Boeing is Q-Tip." In Example 2, each lexical item is replaced by a suffixed bahasa gay variant, so that Lelaki cakep, mau ngésong? becomes Lekes cekes, meses ngeses? (boy cute, [you] want fellate?). Here, the effect is somewhat like an English-speaker substituting for "Cute boy, you'd like to suck him?"the Pig-Latinesque phrase Cutong boyong, wantong sukong 'cute boy, want suck?'. Such linguistic strings, however, are atypical: the language game of bahasa gay is usually played by altering only a single foregrounded word in the utterance, as in 'hungry' in Example 3.
(3) Standard Indonesian: Here lapar is replaced with lapangan, an Indonesian term meaning 'open field'. The result is somewhat like an English speaker saying, "I've been Hungarian for two hours." But this makes bahasa gay rather easy for outsiders to decipher: the meaning of "I've been Hungarian for two hours" soon becomes clear to someone overhearing the phrase. The fact that only one or two lexemes per utterance are typically changed into bahasa gay-often lexemes that do not reveal sensitive information-makes doubtful the argument that it serves primarily as a secret register.
But if bahasa gay so rarely serves the cause of secrecy, why should it exist at all? It appears to act most often to invoke a sense of gay community in a context where many gay men can socialize extensively in civic spaces such as parks, but where they have almost no institutional infrastructure-no places to call their own beyond the corner of a town square, no social recognition beyond the occasional (and often lurid) gossip column. Language here works to stabilize social relations, creating a sense of similarity and shared community. Likewise, many languages in Indonesia, though not Indonesian itself, have honorific registers. The best-known example is Javanese, commonly described as having an overall distinction between High and Low variants (Errington 1985). The relationship between Indonesian and bahasa gay is somewhat parallel to the relationship between High and Low Javanese. For instance, given that High Javanese has a vocabulary of only about 1,000 words (Anderson 1990b), entire utterances in High Javanese are infrequent. Many common terms have no High Javanese equivalent: "the word for table is meja no matter to whom one is speaking" (Geertz 1960:249; see also Agha 1998:162). However, substituting a single High Javanese lexeme in an otherwise Low Javanese utterance marks the entire utterance as High Javanese. Similarly, one or two bahasa gay lexemes move an Indonesian utterance into the register of bahasa gay. The key difference is that honorific registers invoke difference, whereas bahasa gay invokes sameness and belonging.
I encountered one of many instances of this pattern in Makassar. On August 25, 2000, I recorded a discussion meeting held by a group of about 20 gay men at the house of Amir, a well-known community leader who, supported by a local nongovernmental organization, opened up his small house for meetings about the trials and tribulations of the gay world. Such meetings often began stiffly; even though they took place in part of the gay world, their designation as "focus groups" with nongovernmental sponsorship gave them an air of formality worlds apart from Andri's easy banter at Texas, hundreds of miles away on the island of Java. It should not be surprising, then, that although the meeting grew steadily more relaxed and informal, only six widely scattered tokens of bahasa gay appear in the first 30 minutes of the recording. Eventually the conversation moved to the question of interactions at the town square or kampus between gay men, waria, and the normal men who come to kampus in search of sex with one or the other (and who many suspect may secretly be gay men or waria themselves). The conversation grew more and more animated until finally one of Amir's friends told everyone to stop talking and said the following (Example 4):  'We need to be friendly with waria . . . [because with regard to the] problem of these men maybe they really have the attitudes of waria, not the attitudes of men'.
In this utterance, suffixation with vowel shift is established as a bahasa gay game in this conversation. It is here that the conversation takes on a more relaxed and intimate tone and the topic shifts to more personal questions of desire and belonging. Ninety seconds later into the recording, another man commented (Example 5): (5) masalahnya kita memang sémong begitu suka menggoda orang The problem is we who are gay like to seduce people'. Shortly thereafter, the pattern was extended when someone else asked (Example 6): (6) sépong? 'Who?' [sépong = siapa 'who' + -ong] Although this conversation took place on the island of Sulawesi in 2000, several of these terms (including bences and lékong) are in Amen Budiman's (1979) word list of bahasa gay, and all of them would be recognizable to persons conversant in bahasa gay, gay or otherwise, anywhere in Indonesia where bahasa gay has become a presence. To utter only a few words of bahasa gay, then, shapes a larger cultural context. Like High Javanese, bahasa gay marks and structures social relationships. One motivation for the choice of Indonesian as a national language was its lack of registers such as High Javanese. However, at least one register occurs in Indonesian in the form of bahasa gay. This register does not serve the cause of secrecy but reveals and sustains the interlocutor's inclusion in the gay world.
The examples from the Makassar conversation illustrate a third reason why the "secret language" ideology seems insufficient: bahasa gay is usually spoken in the gay world, when outsiders are not immediately present-in a deserted corner of a park, in an apartment, on a bench in a shopping mall. It is rarely spoken in mixed company as a social screen; it typically acts not to distinguish but to include. When this happens (I have heard it used on a bus to comment on an attractive man, for instance), it may temporarily mask the content of what is being said, but such utterances attract rather than deflect attention by their oddity. Also relevant in this regard is that Linda's anger over the disclosure of the "secret" was not semantic but pragmatic, regarding not revealed content but inappropriate use. A similar pattern of use can be found in bahasa prokem; though it is sometimes used as a secret language by criminals and street children, "university students . . . do not use it in public or at home with the intention of not being understood by others . . . they use bahasa prokem . . . among themselves" (Chambert-Loir 1984:116).
This characteristic of bahasa gay is illustrated in the zines (self-published countercultural magazines) published by some gay groups (Boellstorff 2004b). Such zines almost always include short stories about the gay world, sent in by subscribers from across the archipelago. Many of these stories concern life in ostensibly public sites such as parks and town squares that double as locales in the gay world. In this context, determining whether a newcomer is gay can be a significant concern (as seen in the discussion group in Makassar and also in Andri's interactions with Ali at Texas in Surabaya). In fictionalized narratives of encounters in zines, initial contact with other men usually takes place without the use of bahasa gay, and bahasa gay usage indicates that the position of both interlocutors in the gay world has already been established. The following excerpts illustrate two deviations from this pattern. In Example 7, Jim is trying to determine if Dario, a man to whom he is attracted, is gay. The two men are at Dario's home, not in a public place. Dario has just mentioned that he has an MBA, and Jim asks what the acronym stands for: 8 "Just Want Business," Dario joked. "Oh, I thought Just Sex with Young Guys," answered Jim, trying to guess if Dario understood the language often used in gay circles. "My, what does that mean?" Dario asked [. . . ] "Surely you know . . . " said Jim anxiously. "Really, I don't know," answered Dario. "Okay . . . later you'll know. . . it's the language of street youth," said Jim while bringing his body closer to Dario.
(GAYa Betawi 1998:22) The use of bahasa gay by one man to another before it is known that both are gay is usually considered shocking in these zines, particularly in a more public space such as a park or town square, as in the following story (Example 8) from the city of Yogyakarta. A new man has arrived with a friend but is now alone; he approaches the gay protagonist: (8)"Eh, anterin aku yuk!" kalimatnya yang pertama langsung meluncur begitu saja. Wah, ini orang sok akrab banget, pikirku. Padahal aku tidak mengenalnya. "Tadi sama temanku, tapi dia dapat méongan, terus aku ditinggal." Akhirya singkat cerita, aku mengantar orang itu pulang. Kalau kupikir lucu juga, karena kami belum belum saling kenal.
"Hey, take me [home]!" was the first sentence he just threw out at me. Wow, this guy sure pretends to be chummy, I thought. Even though I didn't know him. "I was with my friend, but he got a screw, (méongan 'screw' = méong 'meow' + -an (nominalizer)) so I got left behind." So to make a long story short, I took that guy home. I thought it was funny, because we didn't know each other yet.
(Al Marshal 2001:43-44) In Example 8, the stranger uses the bahasa gay item méong 'screw', a word that means 'meow' in standard Indonesian. The term is doubly motivated because it also appears to have an -ong suffix (although in fact it is a single morpheme). That the use of bahasa gay primarily marks the conversation as too quickly intimate is indicated in that méongan is italicized in the zine itself, as if it is a non-Indonesian term, in the same way that English or Arabic terms are italicized. This is a common technique for the representation of bahasa gay in zines and gives the sense that it is a language in its own right. Once again it appears that bahasa gay works to create a sense of community, even one brought out into the open too quickly, rather than to keep gay community or subjectivity secret.

The Appropriation of Bahasa Gay
Finally, the most important indication that bahasa gay does not act as a secret language is that it is increasingly appropriated by Indonesia's normal world. While friends and family often do not know that someone is gay, gay men can sometimes be openly gay in the presence of normal Indonesians, especially if they work in a salon. These interactions make it possible for bahasa gay terms and even derivational patterns to enter vernacular Indonesian. Bahasa gay thereby becomes part of a national vernacular or bahasa gaul. In the normal world, the register created by switching a word or two in an utterance to bahasa gay/gaul appears to invoke an Indonesian public culture of freedom from official stricture. In recent years the dissemination of bahasa gay has been extended by the entry of bahasa gay terms into mass media. By the mid-1990s, during the twilight years of Soeharto's New Order, gay men commented on how talkshow hosts and celebrity guests on television shows such as Abad 21 '21st Century' or Portret 'Portrait' on the Indosiar and SCTV stations, respectively, would sprinkle their audience patter with bahasa gay terms such as ember (see Table 2) and péres (see Table 5).
Following the loosening of controls on mass media after Soeharto's fall in May 1998, there has been a dramatic rise in the appropriation of bahasa gay. In 1999 GAYa Nusantara, the largest of several informally published gay zines, ran an article titled "Bahasa Gay Menjadi Bahasa Gaul" ("Bahasa Gay Becomes Bahasa Gaul"). The article noted the rapid increase in bahasa gay's presence in the mass media in the late 1990s.
Here bahasa gay is treated as merely a lexicon: Words of the national homo people [kebangsaan 'national' kaum 'people' hémong] . . . frequently slide with ease from the lips of Indra Safera or Eko Patrio [stars of the television program KISS]. . . . [T]heir guests like Anjasmara, Inneke Koesherawati, Hedy Yunus, Cut Tari and others don't fail to speak similar words. . . . It could be said that the program Lenong Rumpi ('Wicked Folk Theater') on station RCTI was the one to begin introducing hémong language . . . [I]t's even said that Jakarta youth that cannot use this language are said to be socially inept and behind the times. [Ibhoed 1999:29-30] Bahasa gay took an even greater leap into the Indonesian public eye with the publication of Kamus Bahasa Gaul 'The Dictionary of Bahasa Gaul' by television personality Debby Sahertian (1999). Once again, the dictionary focuses on terms, not derivational patterns. An instant hit when first published in 1999 and in its 11th edition by 2003, the text openly acknowledges that much of bahasa gaul comes from bahasa gay. In fact, I have heard gay and lesbi Indonesians express frustration at Sahertian for "revealing our secrets," and it is said that Sahertian once apologized for popularizing bahasa gay. Mass media coverage of bahasa gay is also increasing: When a gay man appeared on a local talk show in September 2000 in Surabaya, a newspaper covering the event noted that "when answering questions, he often used terms that are frequently used by the gay community in Surabaya; for instance, 'Texas' for a meeting place and 'endang' for the word 'enak' [good]" (Jawa Pos 2000). More recently, news of bahasa gay's appropriation by Indonesian national culture has reached the international media. In January 2002, the Australian Financial Review ran an article on bahasa gay, which it called bahasa gaul (Dodd 2002). 9

Conclusion
The centrality of difference to understandings of language has been demonstrated by work on language ideology showing the importance of "the ideas with which participants frame their understanding of linguistic varieties and the differences among them, and map those understandings onto people, events, and activities that are significant to them" (Gal and Irvine 1995:970), and by work showing how register "construes differences of speech habit as emblematic of differences in identity, employing language to motivate differences in social identity" (Agha 1998:168). But how can language constitute not only difference but belonging, beyond the mere fact of shared membership in a linguistic community? This is one key question raised by bahasa gay, both in its use in the gay world and in its use in popular culture.
Gay Indonesians might seem to epitomize difference; they seem to lie radically outside the norms of Indonesian societies. Within gay communities and in popular culture, however, bahasa gay appears as a register of belonging, not one of hierarchy or distance. The "social stereotyping" that co-occurs with bahasa gay consistently points toward inclusion in translocal collectivities.
Nowhere do gay Indonesians think the concept gay comes from a Javanese, Balinese, or other ethnolocalized tradition. And nowhere do gay Indonesians think that there are persons outside Indonesia who speak bahasa gay and live lives just like theirs: they are aware that gay and lesbian persons exist outside Indonesia, and a few of them have traveled outside Indonesia or met a gay or lesbian person from another country, but their community is first and foremost a national one. Although bahasa gay is neither necessary nor sufficient for gay subjectivity, it acts within gay communities to concretize individual nodes of the daily gay world, as well as a sense that these nodes are linked in a national network. When normal Indonesians use bahasa gay, they are seen to be hip, not queer; it marks them not as gay but as in tune with popular culture. One possibility is that the national character of bahasa gay can be delinked from its original association with homosexuality because gay subjectivity is so strongly linked to national culture in the first place.
The increasing ease with which bahasa gay has moved from parks and other sites of gay life to Indonesian popular culture suggests it is shifting from a "genre register" linked to context, to a "social register" linked to "stereotypical personality types" (Hervey 1992:198). Its referent is coming to be the user more than the context of use. Bahasa gay can now index two domains of Indonesian life that appear opposed: the world of gay life, still lived largely in secrecy and shadow, and the dominant world of popular culture. What these two worlds share is that they are national worlds. The "stereotyped personality types" invoked by bahasa gay are no longer necessarily homosexual, but they are necessarily national.
The desires of normal Indonesians are understood to operate across differencefemale for male, and male for female. Although transgendered persons such as masculine-to-feminine warias or feminine-to-masculine tombois deviate from gender norms, their desire is understood in normative "heterogenderal" terms (Faderman 1992); warias desire normal men and tombois desire normal women (Blackwood 1998;Boellstorff 2004a). This is what gay (and lesbi) Indonesians have to offer their society; they alone articulate what gay men call a "desire for the same (sama)". When gay men use bahasa gay, their subjectivities "leak" beyond the boundaries of locality and tradition; they find a sense of national community. Similarly, what leaks from bahasa gay as it is appropriated into the national vernacular is a sense of sameness, of shared identity across islands of difference. Bahasa gay sometimes indexes homosexuality, but it registers belonging. (1987:vii). Uncertainty as to what degree of proficiency qualifies one as a speaker makes these statistics speculative.

Notes
6. Independently of my own work, my colleague Dédé Oetomo (2001) has catalogued and analyzed nearly identical derivational processes for bahasa gay, helping to confirm its national character. The discussion in this section thus tracks Oetomo's own typology, except that I have given my own names to the various derivational processes and reordered them in line with the frequency with which I have encountered them in my ethnographic work.
7. In standard Indonesian, infixing of -em-, -el-, and -er-existed historically, but all forms are now nonproductive and appear in such a limited number of words that the variant forms are usually listed separately in dictionaries and "the meaning of the infix is unpredictable" (Sneddon 1996:25). Examples include tunjuk 'point' > telunjuk 'index finger'; suling 'flute' > seruling 'flute'; gigi 'tooth' > gerigi 'serration.' 8. The scene recalls the situation in some Western gay communities, where historically phrases such as "Are you a friend of Dorothy?" could be used to determine someone's sexuality. Few gay Indonesians have MBAs: this story reflects ideals of the perfect partner, not the typical socioeconomic status of gay Indonesians.
9. I thank Sharyn Graham for first bringing this article to my attention.