Digital Technologies, Social Media, Global and Local Languages in Southeast Asia

. This article aligns with the theme, “Overcoming Adversity, Embracing Change: Addressing Challenges in Language and Culture in Asia”, and with the call for papers for this conference. It takes an optimistic but critical stand on questions of digital technologies, including social media in multilingual societies. The telephone, radio, television and more recently the internet were all in their turn (wrongly) seen as heralding the demise of normal face-to-face communication, and as threatening the continued existence of minority indigenous languages in multilingual nations. But the Chinese traditional saying, “A crisis is an opportunity riding a dangerous wind”, remains relevant in the pandemic and post-pandemic era. Technologies are not in themselves language-specific, nor are they necessarily biased towards powerful, global languages. Social media platforms such as WhatsApp, Instagram and others do not force users to shift towards Spanish, Mandarin, Arabic or English. I discuss examples of mixed language use in Southeast Asian social media contexts, taking up the point raised by Deterding (2020, 175) in his article based on his keynote presentation at the previous conference in this series: “perhaps trying to analyse the different languages in Brunei as distinct entities is flawed. In the modern globalised world, languages no longer belong in distinct boxes that can be neatly labelled”. The key argument is that the mixing of local vernacular and powerful global languages does not necessarily signal an impending language shift. On the contrary, such hybrid discoursal practices in social media may be viewed as a minority language maintenance and survival strategy.


Introduction
This article adopts an optimistic but critical stand on questions of digital technologies, including social media in multilingual societies.There has been a dramatic increase in the adoption of digital technologies for the study of languages, literature and cultures, caused in large part by the global COVID-19 pandemic which has forced learning and teaching into blended and fully online modes of delivery.Students and their instructors have had to make rapid adjustments, moving from physical classrooms to reliance on digital online technologies, which include Learning Management Systems (Moodle, Canvas and Blackboard), and other online platforms such as Zoom, MS Teams and Google Meet.Social media platforms, notably WhatsApp, Viber, Telegram and Twitter (now known as X), have taken on additional functions in the learning and teaching both of languages and of content subjects.These adjustments have affected online and social media language choice and use.The article's major focus is on the effects of digital online technologies on languages in multilingual societies such as those found across Southeast Asia.One key issue is whether increased language contact between local, vernacular, minority languages and national, official and global languages will lead to the endangerment and perhaps the extinction of the local languages, many of which are indigenous and only used in small communities.Five more specific focused questions are addressed, although perhaps not fully answered: 1. Do technologies push users towards greater use of "global" languages?2. Do social media platforms push users towards greater use of "global" languages?
3. Do social media platforms promote mixed language use (code-switching/ mixing/translanguaging?) 4. Does social media cause minority / indigenous languages to be marginalised and threatened?
5. Do languages "belong in distinct boxes that can be neatly labelled"?(Deterding 2020, 175) This final question is less specific, but relates closely to the others, raising methodological as well as conceptual issues.
Code mixing is the alternation of two or more languages within the same sentence or utterance.Example 2 is from a Brunei public online discussion forum, The Brunei Subreddit (https://www.reddit.com/r/Brunei/),retrieved on 8th August 2022.
Example 2: Perayaan.Cuti public holiday jatuh hari jumaat/ahad carried forward sehari.It should be noted that not all researchers concur with this distinction between code-switching and code-mixing.Among many theoretical frameworks for codeswitching research, the Matrix-Language Frame model (MLF) of Myers-Scotton (1993;2002) has been one of the most enduring.In this model, one language, often a local or vernacular, serves as matrix language and supplies the grammatical frame, i.e., the function words and structure of the sentence or utterance.Another language, or other languages in trilingual mixing, supplies lexical content words and phrases: nouns, verbs and discourse markers.This is the embedded language.So, in Example 2, the matrix language is Malay, with the noun phrase "public holiday" and the verb phrase "carried forward" in the embedded language, English.
Translanguaging, as defined and developed by Garcia and Wei (2014) and Wei (2018) among others, is proposed as a broader alternative theoretical framework than that of code-switching and code-mixing.Whilst initially it related to interactions between learners and teachers in the classroom domain, it is now applied across all instances of multilingual language use.Translanguaging may therefore be more appropriate to multilingual Southeast Asia.Whilst language alternation and code-switching involve mainly grammatical analysis of spoken and written text, the concept of translanguaging can incorporate notions of identity projection, concealment and negotiation (McLellan 2022;Mendoza 2022).Translanguaging approaches also challenge the whole notion of named languages which are socially constructed through being named (Wei 2018;Wei and Garcia 2022), and in turn cast doubt upon the conceptualisations of bi-and multilingualism, a challenge previously expounded by Makoni and Pennycook's (2007) notions of "disinventing and reconstituting languages".Garcia (2009) challenges standard language ideologies and discourses of national languages which are salient in Singapore, Thailand, Indonesia, Malaysia, Brunei Darussalam and the Philippines.Multilingual users alternate between the powerful national and official languages: Malay, bahasa Indonesia, Standard Thai and Filipino, and the various local vernaculars.This may occur in speech, writing and, increasingly, in keyboarded social media contexts.The notion of translanguaging can help to unravel complex issues concerning how languages and identities are related in the highly multilingual contexts of Southeast Asia. Lee and Wei (2020, 588) provide an alternative definition of translanguaging: "The creative and critical deployment of semiotic resources in communication that transcends normative boundaries between named languages", summarising the term in the constructed word "throwntogetherness" (Lee and Wei 2020, 588).This is found to be especially applicable to social media discourse, both synchronic and diachronic.
However, the suggestion that translanguaging is a replacement for language alternation, code-switching and code mixing, has itself been challenged recently.MacSwan (2022) claims that translanguaging proponents essentialise code-switching research and that their questioning of "named languages" may threaten the linguistic human rights of minority indigenous communities to name and defend their languages as identity markers: "If we take seriously the proposition that discrete languages do not exist, and multilingualism does not exist, …then we cannot meaningfully advocate for children to have access to instruction in their home language" (MacSwan 2022, 29).
This article uses both translanguaging and code-switching approaches, but maintains a necessary degree of scepticism, acknowledging the critique by MacSwan.In the following sections of this article, the five questions listed earlier are addressed.

Do technologies push users towards greater use of "global" languages?
The default interface and rubrics of learning management systems are generally in English, although multilingual platforms can be found: The Learning Lab (https:// www.thelearning-lab.com/blog-elearning-platform/multilingual-elearningplatform)lists 21 available languages; Moodle has language packs available in 122 languages including some indigenous languages such as Te Reo Māori (https:// download.moodle.org/langpack/3.11/).If interfaces and rubrics are predominantly or entirely in English or another global language of wider communication, then less widely used languages may be marginalised and at risk of dying out.
Digital language death (Kornai, 2013) has been recognised for some time as a potential consequence of the widening use of digital technologies, but the findings of Kornai's study are inconclusive as it does not cover EMD through smartphones, which is surely the channel with the greatest potential for maintenance and development of indigenous and minority languages.Among the conclusions of Kornai's (2013) article is the possibility of digital koinés arising through the online mixing of less used languages with more powerful languages with which they coexist: this is happening in Facebook and blog sites such as the public Tutong Kita (Our Tutong, https://www.facebook.com/tutongkita/) in Brunei for the minority Tutong language community (refer Appendix The Bau Bidoyoh language that we use is often mixed with Malay and English.So let us try to use the Bidoyoh as it is spoken in our villages.
Let us share knowledge, share thoughts and spread news at the same time for those who want to learn from those who are more knowledgeable in the language.
(Please respect this group as a place where we can talk and study, and do not use it for spamming, for commercial purposes and disturbing others.The administrator will remove the member from the group without notice whenever this happens.) While the first two paragraphs of this message are in pure unmixed Bau Bidoyoh, the third (in parentheses) contains a number of loanwords from Malay and English, as shown in Table 1.
Purity of language may be impossible to achieve in social media contexts, even when it is the expressed desire of the community.Furthermore, purity may not be a desirable goal.It is quite reasonable to claim that the words listed in Table 1 here belong to Bau Bidoyoh, as they are used by the community of speakers of this variety.These are comparable to French loanwords such as "chef" and "restaurant" which have become part of the lexicon of English, and form part of the inexorable processes of language contact, cross-language influence and transfer.Applying the MLF model of Myers-Scotton (1993;2002) to the third paragraph of the text in Extract 1, Bau Bidoyoh functions as the matrix language, supplying the grammatical framework in both sentences.Malay and English are joint embedded languages which contribute single nouns, verbs and discourse markers.This follows a pattern found elsewhere in mixed language use, worldwide as well as elsewhere in Borneo (Myers-Scotton 2002;McLellan 2009a).The word bidagang here has a Bau Bidoyoh actor-voice prefix bi attached to a Malay noun dagang.It is thus a bi-codal word, containing morphemes from different languages.Mixing can thus occur within words, as has often been found elsewhere, with the grammatical affix from the local matrix language and the root word from one of the embedded languages, e.g., merepair and terstuck, which have Malay prefixes and English roots (McLellan 2009a).
The future of the Tutong and Bidoyoh languages, in spoken, written and keyboarded modes, will inevitably be meshed with the other languages in the repertoire of users of these minority languages, except perhaps on occasions such as traditional rituals and ceremonies, where a purer unmixed variety may be preferred by the community.But any suggestion that mixing of languages may lead to language shift, away from local vernacular languages towards languages of greater power such as Malay and English, needs to be challenged.It is quite possible, as argued by Coluzzi, Riget and Wang (2013) and by McLellan and Jones (2015), that the use of lexemes from the powerful languages in mixed texts serves as a strategy for the maintenance and modernisation of local languages.

Do social media platforms push users towards greater use of "global" languages?
This question connects closely with the first question discussed above.It might be thought that diachronous social media platforms such as WhatsApp, Telegram, Twitter and Facebook Messenger, which permit responses over time, would lead users to choose to communicate in one language only, rather than mixing or translanguaging (Young 2017).This is because they have time to plan and review their texts before posting or sharing them.However, studies of EMD and social media discourse (e.g., Androutsopoulos 2015;McLellan 2005;2009a;2022) have shown that this is very much not the case, and that social media language use closely follows the mixing and alternation patterns of everyday multilingual faceto-face interactions.According to Androutsopoulos (2015, 187-189), networked multilingualism is "a dynamic approach to code-switching online as a resource for the management of social relations and the negotiation of identities".He makes the comparison explicit by explaining that "[i]n order to accomplish pragmatic work that would draw on prosody, language variation or code-switching in ordinary spoken conversation, networked interlocutors manipulate written signs and transcend orthographic boundaries".
Examples from the context of Brunei Darussalam demonstrate that in a corpus of WhatsApp two-party chats and multi-party groups, analysed by 'Aqilah (2019), five basic language choices are available and are used, as exemplified in Example 3.
Example 3: Other examples from a corpus of Brunei Facebook status updates analysed by Nurdiyana and McLellan (2016) show the use of abbreviations and acronyms in both Malay and English for reasons of textual economy, which is an even more salient issue in Twitter, which has a limit of 280 characters per tweet (https:// developer.twitter.com/en/docs/counting-characters).
The emerging pattern is that while social media exerts influence towards the use of global languages, this is very much audience-oriented: Bruneians may post messages in English for an international readership, or other English-knowing Bruneians, but they code mix or use only Malay when their targeted addressees are other Malay-English bilingual Bruneians.

Do social media platforms promote mixed language use (code-switching/ mixing/translanguaging)?
Both of the questions addressed previously are closely connected to this one.A short answer would be "not promote, but permit".Technologies are neutral in terms of language choices.Social media platforms may favour the language(s) of the rubric.
Historically the invention and spread of the telephone, radio, television and more recently the internet have all in their turn been wrongly perceived as heralding the demise of minoritised languages.The same is now claimed concerning the various social media platforms.However, the internet and social media are "unregulated spaces" (Sebba 2009), which makes no prescription for the languages to be used on them.Any language which has an agreed orthographic system can be used on websites and in social media.With minority indigenous languages in Borneo (e.g., Kadazandusun in Sabah, Malaysia), this may involve negotiation within the community as to which of the possible varieties should be preferred and what spellings should be used.Mixed language use is likely to be evident, just as it is in informal as well as formal domains in face-to-face interactions in multilingual societies.
To consider this question in more detail, two further data extracts showing mixed language in social media are analysed.The first is from 'Aqilah (2019, 320), Example 4 is extracted from a group chat between one male and three female interlocutors conducted via WhatsApp, showing the mixing of Brunei Malay and English.
Example 4: (I think mine was left in the UK because it was broken already.)

Key for interlinear glosses: DM = discourse marker; ABBR = abbreviation
There is alternation here between Malay and English in all four of these short messages.In the first turn the abbreviated single English noun "bat" (battery) occurs between two Malay words, and the English conjunction "or" is followed by the Brunei Malay adjective rusak (broken).In the second, the English possessive adjective "my" is sentence-initial, showing English adjective-noun (modifier-head) phrase structure.The third turn has English as its matrix language with switches to Malay in mid-sentence for the two adjectives habis (finished) (Standard Malay, c.f. Brunei Malay abis) and rosak (broken) (Standard Malay, c.f. Brunei Malay rusak).In the fourth turn, both languages contribute to the grammatical frame and to the meaning, thus contravening the MLF model of Myers-Scotton (2002).There is a switch between the first subordinate clause subject "mine" and the Malay verb tinggal.
This exchange of very short WhatsApp messages thus demonstrates patterns of rich intra-sentential mixing which can be attributed to the high level of bilingual proficiency of the interlocutors.This enables them to translanguage seamlessly.
The analysis here distinguishes Brunei Malay, the main in-group variety used by Bruneians, from Standard Malay as used elsewhere in the Malay world.Hence it can be viewed as trilingual, although the percentage of shared cognates between the Brunei and Standard varieties, at around 85%, means that they are mutually intelligible.
The second extract is also trilingual, with English, Filipino and Cebuano.It is from an online Year 8 mathematics class in the Southern Philippines, recorded by Bravo-Sotelo (In preparation), during the time of the COVID-19 pandemic when all learning and teaching had to be conducted in distance mode.Thus it is spoken, not written or keyboarded, but can be nonetheless considered as EMD as defined earlier.
Example 5: English functions as the matrix language here, with embedded insertions in Cebuano and Filipino.There in intra-phrase mixing in turns 1, 3 and 7 by the teacher.When initiating and asking questions in these turns the teacher switches between Filipino, Cebuano, and English, three languages which are shared by her and her students.Students' turns in this extract are very brief and are only in English.Only her turn 5 is in monolingual English with the boundary discourse marker "Okay" followed by an initiation (question) which has three possible answers, "greater than, less than, equal?" Officially, Philippines Department of Education policy stipulates the use of English only for mathematics classes at the Year 8 level (Bravo-Sotelo 2020): by code mixing the teacher is using a subaltern strategy to ensure students' comprehension, especially in the explicit comprehension checks in her turns 3 and 7.In turn 7, the teacher repeats Student 7's correct reply, a standard followup move, then gives a positive evaluation, "Very good", before switching into Filipino to repeat and expand the correct answer, then switching to Cebuano for the comprehension check "Nasabtan?" Similar practices have been found to occur elsewhere in multilingual content subject classrooms across the world (Garcia and Wei 2014).Classroom interaction extracts such as this show how global, national and regional languages, as well as minority languages, can coexist as part of teachers' and students' repertoire in online EMD contexts.
Does social media cause minority/indigenous languages to be marginalised and threatened?
The Internet and subsequently social media were initially seen as channels for the further spread and dominance of powerful national, international and global languages (Young 2017).But these media have long been recognised for their affordances for minority indigenous language maintenance, especially through the rapid spread of smartphones.Another factor here is the more gradual reduction of the digital divide as seen in efforts to improve online connectivity in the rural areas where these languages are used (McLellan and Jones 2015, 27).In 2001 Buszard-Welcher asked "Can the web help save my language?"The example of the Bau Bidoyoh Facebook group (as discussed earlier), which has been in existence since 2009, shows how social media platforms and online technologies can be used to create and maintain spaces for minority language use.Smartphones facilitate the creation of groups through platforms such as WhatsApp, which also enable language maintenance between home-based and diaspora users: often it is those living and working outside their traditional home areas who are activists for maintaining their languages among family and friendship groups, for example Bau Bidoyoh people living and working in the Malaysian capital, Kuala Lumpur, or in the UK, the US and Canada.

Do languages "belong in distinct boxes that can be neatly labelled"?
This question is raised by Deterding (2020, 175) with reference to minority and major languages in Brunei.Deterding suggests that attempts to analyse separately the languages of Brunei are flawed, owing to their being highly interconnected.One example is the closeness of the indigenous minority languages Bisaya and Dusun (Sang Jati).In the earliest cross-language comparative study of the languages of Brunei, Nothofer (1991, 156)  Clearly the translanguaging patterns and tendencies found in social media texts in Brunei, as well as throughout multilingual Southeast Asia, point to languages as leaking and overlapping, not confined to separate and discrete compartments.This was part of the now outdated colonial perceptions of languages, seen in the way that explorers and mapmakers showed language communities living in a fixed space and speaking a named language, which in Borneo and elsewhere was often referred to by an exonym.Examples of exonyms, some of which are still used, are Murut in Brunei and Sarawak (endonym: Lun Bawang), Dusun in Brunei (Sang Jati), Land Dayak in Sarawak (Bidoyoh), and Iban in Sarawak and Brunei (Iban is from a Kayan word hivan meaning "people who migrate").The realities, both historical and contemporary, of migrations, language contact and intermarriage have been downplayed and are only now being uncovered through research in the developing field of migration linguistics (Borlongan 2020).
In Example 4, as in many conversations conducted through social media, there is such rich mixing between Brunei Malay, Standard Malay and English that it is hard to identify a matrix language for the whole conversation, so it is hard to compartmentalise and determine interlocutors' language choice even if we are able to attach the labels of the named languages Brunei Malay, Standard Malay and English.

Conclusion: Connection to Themes of This Special Issue
This article aligns closely with the themes of this special issue of this journal, and with the theme that was framed for the 6th ICLLIC conference, "Overcoming Adversity, Embracing Change: Addressing Challenges in Language and Culture in Asia".In the call for papers, it was stated "Of the many examples of positive changes in human behaviour, one can note the dramatic increase in digital literacy and the rapid adoption of digital technology in everyday life such as teaching and learning".These dramatic changes are social, for example reduction of the urbanrural digital divide, discussed earlier; linguistic, for example new lexis ("lol" is one typical example), and strategic, for example new discourse strategies such as keysmashing (the use of random characters e.g., d6765u/' [-vbc,;dj) to express strong emotions in social media interactions.
The conclusion of this investigation must necessarily be a tentative one, in the light of the rapid development and diversification of social media platforms through apps on smartphones, the likelihood of this development continuing, and the dramatic changes in languaging practices that social media has brought about.
The mixing of local vernacular and powerful national, official and global languages does not necessarily signal impending language shift.There are many examples, both historical and contemporary, of stable bilingualism, in nations which can be described as diglossic, having either high and low languages, or high and low varieties of the same language.On the contrary, hybrid discoursal practices in social media may be viewed as a local indigenous minority language survival and maintenance strategy.As stated by Canagarajah (1999, 75), "The mixing of codes can enable a speech community to reconcile the psychological and socio-cultural tensions it faces between two conflicting languages, and thus maintain both codes".As with the earlier technological innovations, social media is language-neutral and language choices within the diverse social media platforms depend entirely on the users.

Table 1 .
Loanwords in Extract 1 message (third paragraph) used lexicostatistical methods to show that the percentage of shared cognates between Dusun and Bisaya is 82%.Subsequent replication by Muhammad Najib and McLellan (Forthcoming) using the 200-word Swadesh list produced a higher figure, 90%.These figures show that the languages are mutually intelligible, although not identical.Informal interviews with Bisaya community members in 2019 revealed differences in basic terms such as greetings, and a perception of Bisaya as a distinct language.Dusun and Bisaya are both deemed to be dialects of Malay in official government publications, although Nothofer has much lower percentage of shared cognate figures, 41% for Dusun and 43% for Bisaya, with Brunei Malay.A similar point is made byMcLellan (2009b)in a discussion on compartmentalisation of languages, in particular English in the Malay world.