(Trans)languaging, power, and resistance: Bordering as discursive agency

Abstract

The multi/translingual turn in sociolinguistics has highlighted a number of ideological entanglements of foundational concepts, most significantly the way that the notion of ‘named languages’ as bordered entities is intertwined with ideologies of nation and race. In this article, I consider what the conceptual place for linguistic borders is within a ‘trans’ framework of language and propose a focus on bordering, social actions in which indexical meanings at different scales are mobilized to exert control over discursive space by erecting boundaries within or around it. I draw on data from a Facebook group for non-local teachers of English in Thailand, examining how bordering served interests of hegemonic power when linguistic borders were policed with reference to ideologies of nation, as well as how it enabled counter-hegemonic resistance when borders were erected to separate teachers of colour from the intense discursive struggle in the group. (Translanguaging, linguistic borders, bordering, code-switching, scale, resistance, discursive agency)*

Introduction

A defining facet of present thought in sociolinguistics is the critical interrogation of concepts in both academic and popular discourse about language. Much critique has been directed at the notion of language itself, with scholars arguing for a move beyond structuralist conceptualizations of language-as-system and toward a focus on how linguistic resources are used alongside other semiotic resources, and on how their meaningfulness is anchored in specific spatiotemporal moments (see Blommaert 2015; Pennycook 2017; Canagarajah 2018; Li 2018). This movement has been particularly forceful in its calls for the relativization of borders between ‘named languages’ (e.g. English, Spanish, Chinese) and varieties (e.g. British English, Malaysian English), highlighting the historic entanglement of these borders in colonial enterprises, the establishment of nation-states, or both (Makoni & Pennycook 2006; Flores 2013). Such ideologically determined linguistic borders are seen to have grave consequences for the valuation of individual speakers’ linguistic repertoires, since they afford higher levels of linguistic capital (Bourdieu 1991) to particular sets of resources (typically those associated with standard forms of a named language; see Rosa & Flores 2017), by extension pathologizing the linguistic practices of those who do not conform to this narrow, monolingual conceptualization of ‘competence’, legitimating their exclusion from mainstream society.

One of the concepts mobilized in the struggle against this nationalist, colonial, monolingual mindset is translanguaging, defined by Otheguy, García, & Reid (2015:283) as ‘the deployment of a speaker's full linguistic repertoire without regard for watchful adherence to the socially and politically defined boundaries of named (and usually national and state) languages’. Work on translanguaging (see in particular Li 2018) and related concepts (e.g. polylanguaging, plurilingualism, metrolingualism) has sought to establish alternative, practical theories and frameworks for the conceptualization and analysis of language in which boundaries between languages as (imagined) systems are relativized, or even dismissed. The particular focus of translanguaging has been on the tendency to equate ‘named languages’ as categories of social identification (often anchored in ideologies of nationhood) with the structure of the linguistic repertoire of individuals—in other words, that one acquires separate ‘named languages’ as separate competences. This notion of a static, bordered linguistic repertoire is seen in particular to legitimize the already prevalent monolingual bias in social imaginaries of language, particularly in education (e.g. language testing; see Shohamy 2011), and is thus positioned as a key mechanism for the suppression of speakers from minority backgrounds.

The aim of this article is to consider how borders may be re-imagined as a form of translingual practice (Canagarajah 2012). A key starting point is thus that borders have continued relevance to sociolinguistic conceptualizations of language, and that practices in which they are upheld, such as code-switching, are not a priori problematic or discriminatory (cf. the view by García & Kleifgen (2020:557) that ‘describ[ing] the performances of bilinguals as code-switching … uphold[s] named languages as bounded and separate linguistic systems and continue[s] to see bilingual behavior as double monolingualism’). Similarly, ‘named languages’ have continued relevance as resources in metalinguistic awareness and talk (Pennycook & Otsuji 2019; Turner & Lin 2020). What must however be re-imagined is how borders are conceptualized—the conventional view critiqued by the trans-/poly-/pluri-/metro- turn is of linguistic boundaries as static, historic boundaries between systems as naturally distinct as immiscible fluids. Aside from questioning of this bordered ‘totalizing sense of language’ (Stroud 2015:33), a defining feature of the turn has been a greater attention to the role of agency in language, both when it comes to examining agendas underlying the historic invention of ‘named languages’ (Makoni & Pennycook 2006) as well as creativity in use of linguistic resources (Li 2018). A view of borders which takes agency in language—and over language—as its focal point must thus move away from borders as immovable, deterministic structures separating monolingual geographic spaces, instead seeing borders as fleeting, situated discursive edifices, continuously enforced and resisted. The focus then is on actions through which this takes place and the semiotic resources appropriated as mediational means (Scollon & Scollon 2004).

In this article, I conceptualize such actions as bordering, seen as a form of discursive agency in which indexical meanings are strategically employed to draw boundaries and exert interactional power (cf. Khan 2021). I begin with a discussion of why and how linguistic borders are problematized in current literature, using these criticisms as starting points to conceptualize bordering. I then discuss how the relativization of borders in contemporary sociolinguistics aligns with the analysis of online interaction. This reflects the data discussed in the second part of the article, which presents two episodes of interaction collected as part of a study of ideological struggles in a Facebook group for non-local teachers of English in Thailand. Each episode is broken down with regard to participants’ engagement in bordering actions, with particular attention paid to how these were embedded in broader ideological struggles within the group.

From languages to languaging—and bordering

In the first instance, calls to disinvent languages and create a ‘new normal’ for sociolinguistics by embracing the creativity of multilingual speakers stem from injustices which emerged through the establishment of modern nation-states. The chief mechanism for this was the increasing focus on the symbolic role of language that arose with the crumbling of premodern empires, as the legitimation of state power shifted away from the ‘divine right’ of a monarch to rule over their subjects and toward a mandate based on the common identity of members of a particular society (Anderson 2006). Whereas the foundation narrative of most nation-states presumes that this common identity follows naturally from common extraction (Wodak, de Cillia, Reisigl, & Liebhart 2009), there is much evidence of how national identities were planned and policed, particularly with reference to language (e.g. Savski 2018). In general, such identity-building language policies have focused on defining a space for the national body (i.e. setting geographical boundaries of a ‘nation’; see e.g. Winichakul 1997) and on enforcing an idealized form of the ‘national language’ within those boundaries. Through such purification, undesirable linguistic resources and practices have typically come to be excluded, whether their undesirability reflects their imagined foreignness (i.e. belonging to other national bodies; see Spitzmüller 2007) or other types of non-conformity to a unitary national identity (e.g. regional or class belonging, see Winichakul 1997). A good example of this process is the gradual political and linguistic carving up of Southeast Europe, where much language policing has been employed to transform complex, overlapping linguistic continua into a set of bordered languages (Slovene, Croatian, Bosnian, Serbian, Montenegrin, Macedonian, Bulgarian) which, with few exceptions, legitimate the existence of political boundaries. Taking into account that such identity-building also took place against the background of global conquest by European elites, Flores (2013) describes the resultant linguistic regime as ‘nation-state/colonial’.

Though injustices created through linguistic borders have thus been particularly associated with the modern nation-state, they are also a transnational phenomenon. This refers particularly to how linguistic borders which emerged through the nation-state/colonial regime have become part of the cultural flows associated with globalization (Appadurai 1990) and have been appropriated by new regimes at different scales. I refer here specifically to how the boundaries of ‘English’, which may indeed reflect identity-building processes in historic ‘native speaker’ contexts but have since spread far beyond those settings, have been localized in new language ecologies (Pennycook 2010) and have been integrated into new relations of power and inequality (Tupas 2015). ‘English’ has become entangled with instruments of transnational governance, like educational rankings (De Costa, Park, & Wee 2019; Savski 2023) and language tests (Hamid 2014), thus becoming partly detached from specific nation-state contexts and being instead appropriated as an instrument of global corporate power. In postcolonial contexts, the borders of ‘English’ are also entangled in local inequalities of wealth and culture, with elites benefitting from how ideologies of ‘standard language’ and ‘nativeness’ legitimate their hegemonic status. Martin (2014) exemplifies this, arguing that the kinds of inequalities among Englishes traditionally examined by sociolinguists in light of geopolitical relationships can also be observed among different social strata in the Philippines. Similarly, ‘English’ has been mobilized as part of language policy initiatives in Singapore to pathologize the linguistic repertoires of non-elite speakers—narrowed only to ‘native speaker’-like ‘standard’ resources and excluding the widely spoken Singlish (Rubdy 2001).

The embeddedness of ‘English’ in different layers of power and inequality stresses that named languages, as discursive constructs, are not one-dimensional and static edifices, but fluid and dynamic, malleable according to the different ideological agendas that they are mobilized to serve in different interactional moments (Li 2011). This highlights a further, equally key conceptual shift in translanguaging literature, from language-as-system (an imagined edifice which is ‘out there’ for people to instantiate) toward language as dynamic, embodied, and inseparable from action—or languaging (Li 2018, drawing on, among others, the distinction between ‘first-order languaging’ and ‘second-order languages’ made by Thibault 2011; see also an earlier use of ‘languaging’ in Shohamy 2006). From a sociological perspective, a key theoretical contribution of translanguaging is thus the way it has brought attention to the role of agency in language, both in processes through which language is (ab)used as an instrument of oppression and those through which it is mobilized for creative expression (e.g. linguistic innovations; see Li 2018). In taking such a dynamic, agentive vision of something previously conceptualized in a decidedly structural, static manner, translanguaging is in many ways analogous to ongoing theoretical shifts around the notion of ‘culture’, which has traditionally been seen in similar ways as language, as a relatively fixed set of norms somewhere ‘out there’, determining how people feel and act, but generally lying beyond the scope of their agency, being rather imposed by structural factors like history and geography (for a widely used framework of this type, see Hofstede 1991). Just as Street's (1993) notion of ‘culture as a verb’ refocused attention on the fact that the values and practices associated with any culture exist only insofar as they are ‘done’ in a community (see also Scollon, Scollon, & Jones 2012), a ‘language as a verb’ perspective shifts the attention of sociolinguistics toward how languages are ‘done’ in interaction.

It is in this ‘doing’ of languages that bordering is crucial. That is, bordering involves social actions in which indexical meanings are mobilized and manipulated to create boundaries in discursive space, which often involves the construction of hierarchies (e.g. divisions between ‘appropriate’ and ‘inappropriate’ semiotic resources; see Flores & Rosa 2015) or the straightforward exclusion of everything but a narrow set of linguistic resources (e.g. through ‘English-only’ regimes in classrooms). While bordering is anchored in broader social ideologies, and thus draws on cultural resources that transcend any single interactional moment, it is best understood as a scalar process when seen through a ‘language as a verb’ lens. When ‘named languages’ are explicitly invoked through metalinguistic talk (e.g. “Speak English!”) or when individuals deploy a ‘named’ part of their semiotic repertoire (e.g. ‘academic language’), they observe ‘“norms”—orders of indexicality—that are attached to a multitude of centers of authority, local as well as translocal, momentary as well as lasting’ (Blommaert 2007:2, italics in original). Bordering is thus part of a ‘continual process of enregisterment’ (Eckert 2016:72; see also Jaffe 2016) through which indexical meanings emerge and become anchored in social awareness (‘second-order’ and ‘third-order indexicality’; see Silverstein 2003; Johnstone 2013). This takes place at an intersection of multiple flows, reflecting the individual as well as the trans-individual levels—described by Lemke & Lin (2022) as multiple timescales of ‘going along together’, whether in a particular interactional moment or as part of broader, institutionalized frameworks of time (e.g. a class, semester, academic year, study programme).

A conceptual shift from static borders to dynamic bordering also necessitates a rethink of the role of power in the ‘doing’ of languages. Many processes of bordering, particularly those carried out through established, organizational hierarchies (e.g. the nation-state), may appear to be top-down impositions of linguistic differentiation. This may for instance be the case with policies establishing linguistic hierarchies (Savski 2020a) as well as language tests normalizing (or pathologizing) certain linguistic practices (Cushing 2020) and contributing to the exclusion of people from geographic spaces (Khan 2021, 2022). As much work around policy and testing illustrates, however, these instruments are part of discourses which are varied and complex, to the extent that the ideological constructs they seek to enforce are often significantly transformed as they are transferred across scales (Widiawati & Savski 2020). This reflects not only specific ecological differences that exist between sites of, for instance, policy production and implementation, but also broader co-existence and interaction of different voices (heteroglossia) and ideologies (polyphony) in public discourse (Bakhtin 1981; see also Savski 2020b, 2021). Taking account of this pluralistic nature of discourse is key to reimagining borders—rather than being seen as a top-down imposition of boundaries, bordering studies how boundaries are constructed, enforced, and resisted in interactional struggle. This reflects the inherently unstable, fleeting nature of indexical meaning in general—discourses around language are complex and indexical fields are thus rarely anything but heterogeneous, with contestation of indexical links and contradictions in indexical order common themes (see e.g. Johnstone 2013; Jaffe 2016; Jocuns 2022).

Translanguaging and bordering in social media discourse

Much of the research on language on social media and social networking sites has focussed on how these spaces have, from a technological standpoint, allowed new communicative practices to emerge. In particular, existing research has examined the balance between the affordances online spaces offer their users to act creatively and the constraints determining the scope of such creativity (boyd 2010). Such affordances and constraints are linked to the technologies underlying social media spaces and those available to individual users. For instance, social media like Facebook now allow for the use of multiple scripts and embedding of various kinds of visuals (images, GIFs, emoticons), external linking, as well as other kinds of actions like sharing and reacting. However, individual users may also be constrained in their use of such features by technology, for example, needing to Romanize different scripts due to lacking a localized keyboard, as well as by gaps in their own repertoire, that is, not being able to embed GIFs (Androutsopoulos 2015). Additionally, when compared to sites like Twitter, designed as an ‘open space’ where users can see each other's tweets, Facebook is somewhat more closed, given that a (non-celebrity) user's audience is generally limited to their ‘friends’ (Seargeant, Tagg, & Ngampramuan 2012). An exception to this constraint is public Facebook groups like the one examined below, which, while being limited to a set number of users, can become the kind of ‘open space’ offered by Twitter.

A theme in research on social media is their apparent conduciveness to sociolinguistic creativity, in particular to translingual practices. In part, this is linked to technological affordances, as the ability to juxtapose and integrate different scripts encourages the kinds of hybridization observed by, among others, Seargeant & Tagg (2011), Seargeant et al. (2012), Canagarajah & Dovchin (2019), Spilioti (2019), and Li, Tsang, Wong, & Lok (2020). However, another factor is that such spaces have disrupted some of the more rigid social norms surrounding writing. Whereas writing was historically stringently governed by norms of standard language, online writing has relativized much of the traditional rigidity around spelling, punctuation, style, and so on. Similarly, the fact that the monolingualism of conventional writing is often challenged by online writers is perhaps a reflection of how social media spaces tend to bridge the public/private divide, giving users a sense that they are contributing to a less open discourse than with traditional ‘public’ genres like letters to the editor (Seargeant et al. 2012). The lack of institutionalized linguistic gatekeeping, which determines the rigidity of much traditional public writing, can also allow for linguistic norms to be openly challenged online (e.g. Canagarajah & Dovchin 2019; Qi & Zhang 2021), though research on the dynamics of language debates on social media (e.g. Jones 2013) shows that more conventional (prescriptive) ideologies are far from irrelevant in such spaces.

Whether the language practices in particular social media spaces follow conventional norms and enforce linguistic borders associated with the nation-state/colonial regime or whether they become spaces in which such traditional borders are subverted and challenged depends greatly on how these spaces function as nexuses of practice, or semi-stable windows which emerge at the intersection of social practices (Scollon & Scollon 2004). The historical body of a social media space as a nexus of practice, for instance, reflects the life experiences of those acting within it, encompassing their linguistic repertoires and the language ideologies they have been socialized into (Bourdieu's (1991) linguistic habitus). While a social media space can quickly become a translanguaging space in which conventional linguistic borders are subverted (Li 2018), this depends greatly on the extent to which overlaps exist among the linguistic repertoires of its members,¹ as well as on their ideological dispositions toward language. A positive disposition toward linguistic hybridity can emerge as a by-product of the overall dynamics of a social media space, such as the presence of specific discourses in place—different social issues engaged with in the space—or interaction order—situated norms and expectations governing social action (Scollon & Scollon 2004). Li and colleagues (2020) offer an example of how the linguistic creativity involved in creating a Hong Kong news site on Facebook as a translanguaging space went hand-in-hand with a broader agenda of social critique. A similar case is presented by Canagarajah & Dovchin (2019), who examined how Mongolian-Kazakh and Japanese youths used Facebook to develop critical discourse in otherwise restrictive contexts.

The data in context

The research underlying this article took place during 2019–2021 with a principal focus on examining the recruitment of non-local teachers of English from differing countries of origin in Thailand. Such research is highly relevant in Thai education because of the relatively high number of non-local teachers of English active within it (including public and private institutions, primary to tertiary levels) and because of the great diversity and inequality found among these teachers as a professional community (see also Comprendio & Savski 2020; Savski 2021; Savski & Comprendio 2022). While several more affluent Asian nations (e.g. Japan, Korea) have historically imposed limits on the hiring of non-local teachers, barring all but citizens of ‘native English speaker’ nations from employment, Thailand's educational system has no such restriction. Instead, English teachers at Thai schools have traditionally hailed from a variety of countries, with two distinct migratory flows particularly prominent. The first leads from ‘developed’ contexts and consists primarily of white teachers, historically often male retirees seeking to supplement their income by teaching English, but more recently also younger professionals seeking to requalify (Savski & Comprendio 2022). It should be noted that while many of these are nationals of ‘native English speaker’ countries, a significant proportion is not, with citizens of Germany, the Netherlands, Poland, and other non-Anglophone white-majority nations active as English teachers. The second flow, in contrast, leads from ‘developing’ contexts with a GDP lower than Thailand, either other Asian nations, with the Philippines being especially common as a country of origin, or nations in Africa. This flow is particularly instrumental to the spread of English-medium programmes, where qualified, English-speaking subject teachers from postcolonial nations are often a key part of the workforce (Widiawati & Savski 2020).

The distinction between these two flows is relevant not because of geography but as a reflection of the inequalities they are intertwined with in recruitment practices. In general, significant inequalities of both economic and symbolic capital exist, with white teachers (migrating as part of the first flow) idealized and privileged over teachers of colour (second flow) in an often highly racialized discourse (see Savski 2021). Much of this plays out online, particularly on social media where a number of prominent recruitment spaces for English teachers have emerged. My study examined one such space, a public Facebook group for teachers of English in Thailand (with over 40,000 members at the time of writing). I began the research by tracking group activities over an extended period of nine months, during which time I simply observed the dynamics of the discourse among group members, making notes about salient occurrences and getting a sense of the community's historical body. During the observation, it became clear that the group was extremely diverse, reflecting the heterogeneity of the community it targets, but also that significant ideological tensions occurred among members on a daily basis. Reflecting the professional orientation of the group toward English teaching, these tensions revolved both around symbolic issues (e.g. through debates about whether ‘native speakers’ had more authority over English) and the more political-economic ramifications of such symbolic inequalities, related particularly to issues of employment and salary discrimination.

The existence of such continuous tension among group members meant that what I noted in my observation often were instances of direct confrontation, and in particular strategic actions performed by interactants to try and exert power over discourse. Such actions often mobilized indexical meanings at different scales, either implicitly or explicitly, to try and ‘shut down’ an opponent, with several instances typically occurring in the same sequence of comments. My focus on bordering, discussed above, reflects these typical features, as does the focus of the following section around two interactional episodes in the group. The benefit of examining episodes in analysing bordering is that it allows for a more nuanced examination of the scalar nature of how indexical meanings are mobilized, since actions can be examined against the background of an extended interaction, with reference to previous discourse in the group and its historical body, to the habitus of individuals (to the extent that it may be observed through interaction), to resources like language ideologies, as well as to the set of political-economic relations the interactions are set in. Thus, my discussion of both episodes is to be read in the spirit of moment analysis, in which a focus on identifying patterns and regularity is relativized in favour of a more in-the-moment examination of individual (inter)actions, though without dispensing with examination of broader discourses into which such (inter)actions fall (Li 2011). In both episodes, I examine how individual group members mobilized indexical meanings when engaging in bordering actions, as well as considering how these actions enabled (or hoped to enable) them to exert control over group interaction. The episodes occurred within a period of five days at the end of the nine-month observation period described above, the first within a single comment thread and the second in two closely interrelated threads.

Two episodes of bordering

Episode 1: Bordering, upscaling, and suppression

The first episode occurred in comments below the following post.

Dear fellow Filipinos,

Please be informed that we, Filipinos, aren't Native English Speakers(NES). So whenever we see schools that demand NES only as their staff, better skip the ad and look for a job hiring which welcomes Filipino applicants to apply. Just a friendly advice. Goodluck to all of us, mga Kabayan!

The post reflects a broader tension around the status of teachers from the Philippines in Thailand. While many Filipino teachers are active, often fulfilling key roles in Thai schools (e.g. teaching in English-medium programmes), their economic valuation is typically lower than that of white teachers, regardless of nationality or qualification. Indeed, most jobs offered in the group were targeted specifically at ‘native speakers’, or used various kinds of racialized language (e.g. ‘native speaker or European’) to exclude teachers of colour like Filipinos (see Savski 2021). There was much debate around this inequality in the group, with some members attempting to legitimate or enforce it and others attempting to challenge or subvert it. The interaction below this post in (1) is a good example of this dynamic.

The exchange begins with a resistant move by a Filipino teacher (line 1), and continues with a sequence of attempts to suppress this counter-hegemonic discourse (lines 2–4). The same sequence then repeats, with an act of resistance (line 5) countered again (line 6), followed by a retort (line 7). It is significant here that those seeking to legitimate the unequal status quo and suppress resistance are doing so by strategically constructing linguistic borders and excluding resistant actors from them. In line 2, bordering is done explicitly through metalinguistic talk, with a reference to ‘better English’, whereas in lines 3, 4, and 6 the members engage in policing the borders of this code by highlighting unacceptable elements of spelling and syntax.² These bordering actions are examples of upscaling (Blommaert 2007) in that they strategically appeal to a higher centre of authority—in this case, the timeless ‘named language’ with set rules and a clearly delineated ‘ownership’. Indeed, what is playing out through these acts is the legitimation of a much broader inequality between speakers of English, not specific to teaching English in Thailand but anchored in a centuries-long process of colonization and imperialism (Phillipson 1992).

Though such strategic upscaling invokes a historicized, static vision of language as rule-governed, there is in fact a much more complex, layered process of bordering at play. The borders being erected are very much relative to the affordances and constraints of Facebook as an interactional space, considering the various semiotic resources being deployed, including annotated embedded screenshots (see images in lines 4 and 6). These transcend the traditional borders of ‘English’, yet there is little enforcement of these boundaries. Furthermore, the bordering is embedded in the historical body and interaction order of the group. Much of the interaction appears to be conducted in a particular key, best described as ‘edgy humour’ (i.e. humour which tests boundaries of social appropriateness). My observation of group interactions suggested that this key was a central part of the interaction order of this community, with features like sarcasm (line 4), the attribution of homosexuality to the interlocutor (line 5), and the use of profanity (line 7) common in interaction, particularly when discussion turned to controversial issues like inequality and tensions began to appear. This points to a slight disconnect between the idea of ‘English’ being invoked through appeals to language ideology and the semiotic practices in the community—while explicit talk centres on the idea of purity and adherence to a translocal, timeless ‘named language’, how ‘English’ is done in these interactions is a different story, and underlines the rather selective, strategic nature of bordering.

The episode continues within the same comment thread in (2).

While at first sight this appears to be a repeat of the sequence above, with an attempt at critiquing hegemonic inequality (line 1) rebuked through language policing (lines 2–3), the rest of the interaction suggests this is in fact a much more complex case of meaning construction. The Filipino member who initiates this interaction (A) had previously been involved in interactions and had expressed support for the preferential treatment of ‘native speakers’ as well as others with higher language proficiency, placing themself at odds with many other Filipino teachers in the group. Here, the user reinforces this position by performing a ‘mock code’, a highly stereotyped representation of ‘Filipino English’ achieved by combining a series of grammatical deviations (non-standard uses of prepositions, spelling, aspect, word order) into a single utterance (line 1). This attempted performance of a mock code is, however, rejected by another member (line 2), to which A responds with a switch back to standard English when delivering their rebuttal (lines 3–4). Despite this switch, perhaps motivated by A's wish to display competence in standard English, the legitimacy of the mock ‘Filipino English’ is again rejected (line 5). I should remark here that, considering the ‘edgy humour’ key characteristic of this group (see above), these rejections may also be read as mock moves, since the three users may well have been aware of the intent behind the original post but may have rejected the move in order to gain a rhetorical advantage. Another user, also a Filipino, does acknowledge the intention behind the post and offers a similar construction (line 6), receiving a positive response from the original author (line 7) in a move which is also notable for making use of a resource highly specific to the Thai context, the use of the repeated numeral 5 to signify laughter (relying on the Thai pronunciation of this numeral, /hâː/).

This interaction represents another instance of strategic bordering. Whereas in the previous example this involved explicit references to an imagined ‘good English’, the route taken here is more complex. In particular, the interaction offers an instance of crossing—‘use of a language or variety that, in one way or another, feels anomalously “other”’ (Rampton 1999:54). This ‘otherness’ of the language is explicitly signalled when A later describes line 1 as an imitation of ‘not being proficient in English’ (line 3) and ‘grammatically incorrect sentences’ produced ‘on purpose’ (line 4). A draws on a broader set of discourses in place (attempting to invoke a stereotyped image of Filipino English) in order to dissociate themself from others in the interaction, seemingly to accrue added linguistic capital by assuming a level of authority closer to ‘native speakers’. This aim is further supported in line 4 by the member's claim to knowledge of ‘western culture’. B and C, both ‘native speakers’, appear to refuse to acknowledge the legitimacy of this crossing, illustrating Rampton's (1999, 2009) observations about how such moves often do not pass without commentary. This struggle over interactional positioning puts focus on the differentiated historical body of the group—whereas this instance of crossing appeared to be meaningful as a mechanism of self-othering to another Filipino (D), socialized into the complex discourse around language and social structure in the Philippines (see e.g. Martin 2014), it was not acknowledged by two ‘native speakers’ in the interaction, either due to their unfamiliarity with the indexical field being invoked (i.e. one in which a ‘Filipino’ identity can be indexed by standard English) or due to their unwillingness to accept A's claim to authority, or both.

Considering this episode, a few broader points can be made about how and why bordering involves strategic mobilization of indexical meaning. It appears clear from the exchanges that a key function of bordering was to strategically silence resistant members of the group. This is evident in how ‘good English’ was policed selectively, with comments like lines 2, 4, and 6 in (1) and line 2 in (2), making rather pedantic references to ‘rules’ only when such upscaling was useful to strategically silence others by invoking a higher, more generic centre of authority. That is, the ideologies ‘English’ as a named language brings with it, particularly those associated with the nation-state/colonial regime and the assumed global ownership by ‘native English speakers’, were relevant, but often only insofar as they were made relevant by agentive moves of individual group members. The aim of this agency, considering the historical body of the group and the ongoing power struggles within it, was to exert control over discourse. Indeed, considering the dynamics of both interactions, the bordering seen here did not so much involve separating ‘English’ from ‘non-English’, or ‘good English’ from ‘bad English’, but rather sought to lend legitimacy to the voices of those in power and delegitimize those critiquing power. The action of crossing into a mock ‘bad Filipino English’ in line 1 appeared to aim at overcoming the silencing effect of this bordering on Filipino group members, though this was not entirely successful.

Episode 2: Bordering, translanguaging, and resistance

The second episode occurred within five days of the first. It was sparked by a post from a member who had also participated in the first episode, consistently attempting to suppress resistance and enforce the status quo (see comments in lines 4 and 6 in (1) and line 5 in (2)).

A: If people keep posting Tagalog on this forum, I'm going to start using three syllable words. Fair's fair!

The post, attempting to enforce an English-only language policy in group discussions, came after several instances of Filipino members using Tagalog. It was phrased in the ‘edgy humour’ key characteristic of the group (see above), flirting with offensiveness to Filipino members (implying that three syllable words are as incomprehensible to them as Tagalog to the ‘native speaker’ writer). It garnered a number of replies, many written in a similar ‘edgy humour’ key but others also departing from it.

This sequence begins with much the same kind of interaction observed above, with two challenges to the anti-Filipino sentiment of the original post (lines 1 and 3) suppressed through strategic bordering (lines 2 and 4). The key described above also continues to be relevant, with posts continuing to make use of various degrees of mock insults (‘crazy old man’ in line 1, ‘bald old man’ in line 3). A shift occurs when the author of the original post switches to Tagalog (line 5). This is a significant event not only because it constitutes a move away from the monolingual interaction up to that point, but also because the writer, a British ‘expat’ and one of the most active members of the community, had never previously made use of any language other than English. Given that this was the only instance of this individual posting in Tagalog, the most likely explanation is that the writer had used the affordances of online writing—the availability of automatic translation services—to prosthetically expand their repertoire, and that this is thus a largely performative, emblematic use of Tagalog (Poplack 1980). The meaningfulness of this switch is also unconventional: considering that the same member had attempted to ban the use of Tagalog specifically, it can be assumed that the purpose was not to dissent against English monolingualism, but rather to enforce it by momentarily appropriating Tagalog. In other words, compared to the crossing attempted in line 1 in (2), this does not appear as an attempt at ‘self-othering’, but rather at maintaining the otherness of Filipinos. It can, however, still be conceptualized from a crossing perspective, particularly because of the responses it evokes from Filipino group members (see (3), lines 6–10). Among these, two involve code-switches, one in line 8 into Ilocano—seemingly also a departure from the otherwise dominant ‘edgy humour’ key—and another in line 10 into Cebuano.

This extract exemplifies how bordering involves agentive manipulation of indexical meaning. A's contributions to the interaction appear to invoke a key ideological construct of the nation-state/colonial regime, namely the association of a particular national community (Filipinos) with a single named language (Tagalog). Indeed, a sign of the relevance of national identity is the fact that one of the comments (line 9) explicitly engages in talk about Filipino identity, subverting A's contribution through a reference on how ‘we’ (Filipinos) are not only ‘sensitive’ but also ‘sentimental’ and ‘drama queen’. In this vein, the overt displays of multilingualism in lines 8 and 10 can be seen as resisting the essentialized indexical link that A attempts to impose, instead associating Filipino identity with a much broader repertoire. Two named languages, Ilocano and Cebuano, are explicitly invoked by group members as a means of resistance to attempted enforcement of an English-only policy—and by proxy to the unequal status quo that this monolingualism legitimizes. The observation that can be made here is that, while bordering can be employed as a means of silencing, it can also serve as a resistant strategy. By introducing Ilocano and Cebuano into the interaction, C and E not only resisted A's appropriation of Tagalog and a reductionist portrayal of their repertoires, they also erected a boundary in the discursive space where this interaction takes place, since this removed A's ability to prosthetically extend their linguistic repertoire—major automatic translation services do not offer Ilocano, and while Cebuano is offered by some, the profanity used here was not translatable at the time of writing (for discussion of how the availability of automatic translation services impacts language use on Facebook, see Hendus 2015).

The use of bordering as a resistant strategy continued to be relevant in the continuation of this episode, which took place in a thread under a job advertisement posted shortly after the interaction above.

This interaction starts with a comment from A, the author of the job advertisement and a white ‘native speaker’ teacher who references the discussion in the previous thread and complains about ‘shit posters’ (line 1)—this reference is not completely clear and may refer either to comments that were deleted before the data was collected, or to private messages. From this point on, the interaction becomes increasingly more complex, as a sarcastic post by a Filipino teacher in Ilongo (line 2) is followed by another instance of a non-Filipino using automatic translation tools to post in Tagalog (line 3), followed by a highly offensive post which seems to liken Ilongo and Tagalog to non-human gibberish through the use of a meme (see image in line 4). From this point on, the interaction changes, being predominantly conducted in Tagalog at first (lines 5–7) before transitioning to mostly Ilongo (lines 8–18). While these languages are prominent in this part of the interaction, it is worth noting that a number of elements, largely individual words or phrases, that may at least to some level be associated with English (e.g. ‘double check’, ‘english’, ‘peace’, ‘losers’, ‘assholes’), are also made use of.

This interaction highlights the need to move away from a conventional view of linguistic borders as determined primarily by the existence of distinct linguistic systems. While a shift could be argued to take place from mainly Tagalog to mainly Ilongo (lines 7–8), this does not appear to reflect any of the meanings traditionally associated with code-switching (e.g. topic change, addressee specification; see e.g. Gumperz 1977) and instead the entire interaction from line 5 to line 18 seems to be a case of languaging ‘without regard for watchful adherence to the socially and politically defined boundaries of named (and usually national and state) languages’ (Otheguy et al. 2015:283). Similarly, the use of isolated ‘English’ elements throughout is difficult to see as conventional instances of code-switching or code-mixing, considering again that these are used among interactants with overlapping language repertoires, particularly since many of the words and phrases being used here may well be seen as instances of the kind of fluid interplay of linguistic repertoires that translanguaging focuses on (Li 2018). Uncritically imposing a set of pre-determined ‘named languages’ on this interaction would almost certainly lead to a misrepresentation of the intertwined linguistic repertoires of the authors, and consequently to their delegitimization as epistemic subjects (Phyak & Sah 2022).

However, this interaction also highlights the need for a focus on bordering as an agentive process. My interpretation of what takes place in (3) in lines 8 and 10 and particularly in (4) in lines 5–18 is that the participants, when faced by direct attacks from group members attempting to enforce a type of elite closure (marginalizing resistant actors by ‘employ[ing] official language policies and … nonformalized language usage patterns’, Myers-Scotton 1993:149), collaboratively established ad hoc translanguaging spaces in which boundaries between ‘named languages’ could be broken down (Li 2018) and, most crucially, in which they could interact with freedom, without being exposed to aggression. The key reason why the participants could establish this alterative interactional space was their collective, strategic bordering, switching away from English in (3) in lines 8 and 10 and particularly in (4) in line 2 and onward, since this allowed them to create separate ‘safe’ discursive spaces from which others, namely those seeking to silence them, were excluded (for a similar case, see Capstick & Ateek 2021). This allowed the interactants to engage in resistant discourse without outside intervention from non-Filipino group members. In (3), lines 8 and 10, they engaged in language play (Qi & Zhang 2021), creatively trading insults in different languages. In (4), the participants reflect more comprehensively on the injustice of the status quo (lines 5–7), forge ties of solidarity (lines 8–9), reflect on conditions in other contexts (line 13) and construct a shared resistant voice by jointly demystifying the ideologized image of the ‘native speaker’ (lines 16–18). Compared to the events of the first episode, where resistance was consistently suppressed before meaningful critique could develop, the interactions in the second episode display a more profound, emancipatory nature, underlining how translanguaging, as a transgressive practice in which conventional linguistic borders are relativized or flouted, can mediate resistance to broader ideological hegemony (Li et al. 2020). However, this episode also provides an account of how bordering can serve emancipatory purposes when mobilized to provide interlocutors with a discursive space in which they are at least partly removed from ideological struggles like those continuously observed in this group.

Conclusion: Bordering as (re)structuring of discursive space

This article focussed on imagining what role borders can have in sociolinguistic analysis at a time when trans-/pluri- frameworks of language have problematized assumptions regarding the natural separateness of ‘named languages’. In particular, these approaches highlight the ideological underpinnings of how languages are ‘named’ (identified, distinguished, legitimized) by focussing on the political nature of such ‘naming’ (Otheguy et al. 2015), critiquing especially how the structuralist conceptualization of languages as naturally distinct systems has invariably ended up being conjoined with ideologies of nationhood, colonialism, and market capitalism (Flores 2013). This article has argued for a re-imagining of borders as bordering, understood as agentive manipulation of indexical meaning—that is, as something we do through our languaging rather than something that is in language. Drawing on data collected from debates in a Facebook group for non-local teachers of English in Thailand, I focussed in particular on the scalar nature of bordering, as a process which engages indexical meanings associated not just with ‘grand narrative’ ideologies like nationalism but also with the affordances of online writing, the interaction order, and historical body of the community, as well as one reflecting the strategic aims of individuals in interaction (e.g. silencing opponents, building resistance). I highlighted how bordering serves both interests of those in positions of power, particularly those co-opting hegemonic ideologies of nationhood and nativeness to legitimate their own privilege, as well as those seeking to challenge this privilege.

A holistic view that can be drawn on the basis of this study and of other work in applied linguistics and sociolinguistics is that bordering constitutes a type of (re)structuring of discursive space. That is, linguistic borders are often drawn and redrawn as a means of enforcing ideas about who, when, where, and how they may participate in public discourses, what identities they may express, and what beliefs they may voice. As translanguaging scholarship has highlighted, a key effect of the imposition of monolingual ideologies in educational spaces is the exclusion of those who do not conform to a particular imagined identity, typically drawn along lines of race (Rosa & Flores 2017) and class (Cushing 2020). As Shohamy (2011) and Khan (2021) show, the same is true when considering how language tests are used to delegitimize the repertoires of immigrants and justify their continued exclusion. The dynamics of the Facebook group examined in this article appear in many ways to be a microcosm of such broader tensions, particularly considering how often bordering actions appeared to be strategically performed to try and silence opponents, especially those seeking to speak out against the status quo. These events underline the fundamental association between languaging and discourse, that is, between meaning-making of any kind and the reproduction of relations of social power (Fairclough 2001), as well as the resultant impossibility to study language without engaging with power in some regard.

A final observation may be made on the way that bordering as (re)structuring of discursive space facilitated subversion and resistance in the group. ‘Grand narrative’ language ideologies like nationalism and native-speakerism can easily appear all-powerful, their constructs so defining to the collective ‘common sense’ that there is little prospect of grass-roots critique emerging, much less that any reconfiguration of underlying political-economic relations may take place. One look at the discourse of the particular online community mentioned above stresses the dynamic, continuously unstable and renewing nature of any hegemony in the face of resistant action. Many of those seen above were instances of linguistic citizenship, actions which seek to ‘[reframe] semiotic practices of citizenship away from a totalizing sense of language’ (Stroud 2015:33), that is, to disrupt entrenched associations of language and redraw lines of indexical meaning. Bordering appeared central to this linguistic citizenship, not as enforcement of top-down borders but as their reconfiguration in development of ‘alternative forms of being-together-in-difference’ (Stroud 2015:34), facilitated by the establishment of new discursive spaces from which more oppressive voices were excluded. This highlights an important final point about borders: at a time of widespread orientation to production of new concepts in sociolinguistics, it is key, borrowing Stroud's words, not to replace a bordered ‘totalizing sense of language’ with a borderless one, but to examine how linguistic citizenship in different ecologies constructs multiple ‘alternative forms of being-together-in-difference’.