A ‘new normal’ of code-switching: Covid-19, the Indonesian media and language change

The COVID-19 pandemic has changed behavioural norms and how people conceptualise everyday life. It has led to prolific use of specific terminology that is new or was previously outside the lexical boundaries of common use. Terms like ‘social distancing’, ‘lockdown’ and ‘new normal' were previously jargon limited to specialist fields. The COVID-19 pandemic which spread globally in 2020 has led to great social change and an associated lexical influence. To study this phenomenon, we examine the lexical effects of COVID-19 on the Indonesian language, through analysis of two well-known Indonesian national newspapers – Kompas and Suara Pembaruan, for the month of May 2020. This was at a time of growing awareness of COVID-19 in Indonesia, that included a partial lockdown in Jakarta. As such, there was a great deal of attention to COVID-19 in the mass media. To study this, we apply quantitative content analysis to the sample data to identify the range and frequency of words borrowed from English. We examine this use of code-switching to also undertake qualitative analysis, exploring the various socio-linguistic dimensions of those borrowed terms. Some usage was found to address lexical gaps in Indonesian language, where other usage appeared more for stylistic, emphatic purposes, drawing on the semiotic power of English in the Indonesian context. Code-switching reiteration was particularly prominent in the sample data. We argue that through code-switching reiteration, the print media can introduce new foreign vocabulary to Indonesian readers, which subsequently generates opportunities for language change. COVID-19 has expedited this process, meaning that there has been an increased likelihood of Indonesian language change during 2020.


INTRODUCTION
It is a rare linguistic phenomenon to observe such a rapid global spread of certain terminology as has occurred with the COVID-19 pandemic. As it spread throughout the world it has changed the way we live and generated a new vocabulary bank that influenced languages on a global scale. According to Oxford Languages (2020, p. 6), 'in almost realtime, lexicographers were able to monitor and analyse seismic shifts in language data and precipitous frequency rises in new coinages' due to and so published a 38-page report detailing the phenomenal breadth of language change that occurred in 2020.
In a short time frame English terminology related to medicine, epidemiology and social change dominated world headlines and permeated different languages around the world, including Indonesian. This lexical expansion predominantly occurred through the process of code-switching (CS), where English words were readily adopted into Indonesian. Recognising that 'great social change brings great lexical change' (Oxford English , the COVID-19 pandemic has shown social factors rather than linguistic factors can trigger codeswitching (The Linguistics Encyclopedia, 2002, p. 70). Social factors include new norms of behaviour, with culturally unpalatable concepts of social isolation and social distancing implemented through lock downs. Much of this was conceptualised and subsequently disseminated in the English language, through the World Health Organisation (WHO).
In practical terms, 'global public communication is conducted only through a small number of the world's languages' (Piller et. al, 2020, p. 504) where communication structures dealing with the pandemic 'replicates this linguistic hierarchy ' (p. 505). The WHO has promulgated official information on the pandemic predominantly in English, but information is also available in other official languages of the United Nations of Arabic, Chinese, French, Russian, and Spanish (WHO, 2020). It has then been up to individual member states, such as Indonesia, to interpret and translate relevant WHO COVID-19 related information through local national health authorities and media outlets (Piller et. al, 2020, p. 505). This dissemination process has created an avenue for English pandemic related terminology, acronyms, and phrases to enter the Indonesian language.
Since independence in 1945, contemporary Indonesian language has increasingly reflected an English influence, most obviously through loanwords. Historically, Indonesian has been receptive to the influence of foreign words and is considered 'a sponge, readily absorbing elements of other languages and integrating them into one language' (Djojonegoro, 1995, p. 6 cited in Holland, 2010. This is because foreign adopted words 'can easily be transformed into various parts of speech via Indonesian affixation once the words have undergone phonological and morphological nativization' (Sugiharto, 2009). Whilst the adoption of loanwords into Indonesian is not a new phenomenon, we argue that the COVID-19 pandemic has greatly accelerated this process -with much of that change being facilitated by the media. In drawing from empirical evidence from print media, we demonstrate how CS reiteration has been used to introduce new concepts. From this, we argue that CS reiteration in the print media is a key driver of language change in Indonesian.

English code-switching in Indonesian
There is a considerable body of literature about code-switching (CS) as a common socio-linguistic phenomenon in multilingual communities where contact between different languages occurs (Gardner-Chloros, 2009, p. 4). As a useful definition, Poplack (1980, p. 583) defines CS as the 'alternation of two languages within a single discourse, sentence or constituent'. Taking an analytical next step, Gardner-Chloros (2009, p. 20 & 23) considers CS as the genesis of language change because CS forms 'one of the principal mechanisms of borrowing'. In examining the nature of CS, Mahootian (2005, p. 362) considers that 'in some contexts code-switching is itself an instantiation of language change'. CS reiteration where the foreign word is repeated in the primary language as a direct translation is what Harjunpää and Mäkilähde (2016, p. 163) suggest is at the 'intersection of codeswitching and translation'. Sneddon (2003, p. 191) describes this phenomenon as a 'loan translation' or a calaque and asserts that within the Indonesian context, 'a great many Indonesian phrases employ native words based on English models' and 'many are phrases that are direct translations of the English'.
There are several possible reasons why English has been code-switched into Indonesian since COVID-19 has emerged. First, English CS may be employed to fill a lexical gap because Indonesian may not possess sufficiently precise equivalent lexicon to express new English pandemic related terminology. To address this gap, English is used for foreign concepts that are presumably unfamiliar to Indonesian readers. In this case the English can be seen 'to fulfil a need for a particular concept' which is new to Indonesian society or which is not 'fully or precisely covered by existing words' in Indonesian (Sneddon, 2003, p. 186). This is certainly the case for newly created concepts and terminology related to the COVID-19 pandemic. Before the pandemic, words such as 'new normal', 'social distancing' and 'lockdown' were largely unfamiliar to the world but are now part of a global lexicon related to the pandemic.
The key focus of this study is CS reiteration, that is where the English code-switched term is repeated in Indonesian. We found this was frequently used in the sample data. Differing from 'basic' CS, CS reiteration is used to emphasise, reinforce or clarify the foreign loanword (Eldridge, 1996). CS reiteration signals the use of a foreign concept before providing a paraphrased equivalent. English CS reiteration in this manner can be regarded as having a communicative purpose, because if Indonesians 'are not sufficiently familiar with the English terms, then providing an Indonesian equivalent, by way of paraphrased translation, may act as an effective communicative strategy' (Foster & Welsh, 2017, p. 403).
Even if suitable Indonesian terms exist to describe pandemic concepts, they may not be used for various reasons relating to practicalities of language production and stylistic purposes. One reason seems to be that a great deal of pandemic news originates from English speaking countries or the WHO. Sneddon (2003, p. 177) suggests that in accessing foreign news, Indonesian journalists 'frequently do not have time to look for appropriate Indonesian translations' to meet the workplace deadlines and commonly use CS reiteration with an 'Indonesian word or phrase to be followed in parentheses by the English original ' (p. 179). This is an important consideration for this study as we draw on data from the Indonesian media.
Another important reason for the use of CS reiteration is for stylistic purposes, where the reiterated English words may not appear necessary for communicative purposes. The use of English CS through reiteration where an Indonesian word already exists may seem to be redundant in communicative terms, but its purpose is likely to be emphatic (Eldridge, 1996). Drawing on the semiotic power of English as a language of prestige and high status to convey an authoritative tone or sense of credibility, CS is not merely for linguistic purposes but is also reflective of social and power relations (The Linguistics Encyclopedia, 2002, p. 71).
It is through these various methods of CS that the mass media has a considerable influence on the way English language is presented to the Indonesian public and may consequently change the Indonesian language (Sneddon, 2003, p. 177). The important role that mass media plays in the dissemination of language is recognised by language planning authorities in Indonesia who have advocated for Indonesian language editors of every newspaper to ensure the correct use of grammar, vocabulary and terminology is being printed (Sneddon, 2003, p. 150). This reflects what Labov (1994, p. 78 cited in Milroy, 2001 identifies as linguistic 'changes from above', where words 'are introduced by the dominant social class, often with full public awareness'. To examine the prevalence of English related pandemic terminology through code-switching this study draws data from two Indonesian newspapers -Kompas and Suara Pembaruan. The use of English CS in Indonesian newspapers is not a new phenomenon. In 1999, the New York Times highlighted that Kompas 'often inserts English words in brackets after Indonesian terms' (Fuller, 1999). However, this phenomenon has accelerated as a result of the COVID-19 pandemic where there has been a rapid influx of English related pandemic terminology. This study draws into focus tensions between the mass print media and the Indonesian Government's Language Centre of the Ministry of Education and Culture [Badan Bahasa Kementerian Pendidikan dan Kebudayaan], hereon referred to as Badan Bahasa [language agency]. These tensions reflect deeper ideological divisions between a conservative view that values a standard form of the national language, as opposed to a more progressive view, that sees a language as needing to linguistically interact with other languages to respond to dynamic contemporary socio-cultural factors. The focus here is on the borrowing of English words in the Indonesian language mass media in responding to the COVID-19 pandemic, and what this suggests about the complex space of global linguistic dynamics.

METHOD Researching English code-switching in the Indonesian media
To explore contemporary language use arising from the COVID-19 pandemic, we use Indonesian news media as the site of research. Stokes (2003, p. 54) argues that media texts are worthy of research because they are 'socially relevant' and constitute 'part of our world', as 'social phenomena'. As such, analysis of media texts offers insights into cultural dimensions and contemporary matters. Moreover, media texts contain language that has been carefully considered and undergone the scrutiny of editing and approval (Mahootian, 2005). Media texts are enduring records that document deliberative acts of language production. Globally and locally, the mass media remains influential, not only in terms of informational content but also in terms of linguistic form. Indeed, the mass media offers great potential for modelling and analysing linguistic acts and the language it produces. In the Indonesian context, Sneddon (2003, p. 177) notes that the 'mass media' plays a significant role 'in exposing the public to English'. In short, researching media texts offers considerable potential to analyse contemporary language use.
For this study, sample data is drawn from two Indonesian national newspapers -Suara Pembaruan and Kompas. From the current 172 Indonesian newspapers, Kompas and Suara Pembaruan have consistently been recognised by the Badan Bahasa as being among the best national print media in terms of Indonesian language use. Kompas has received special recognition for its dedication to Indonesian language, based on the use of spelling, word form and word selection (Aksan, 2019).
Generally regarded as Indonesia's most prestigious newspaper, Kompas was ranked fifth most popular newspaper worldwide in 2019 (4 International Media & Newspapers, 2019). An independent media research organisation, Mediabuzz (2020), has ranked Kompas as number 1 in Indonesia while Suara Pembaruan is ranked at number ten. These newspapers are quite different in format and presentation; Kompas is a broadsheet publication with longer articles, whereas Suara Pembaruan is more of a tabloid with shorter articles, as reflected by the comparative word counts of the sample data for each newspaper in Table 2. These inherent differences are useful in diversifying language sampling, with Kompas symbolising the upper echelons of Indonesian journalism and Suara Pembaruan being in the upper rankings of a more mid-ranged newspaper. Daily circulation of Kompas is around 469,298 with an average readership of 1,409,784 with each copy being read by 2-3 persons (Kompas, 2020). Moreover, Kompas readership is thought to have increased by 30 percent during the pandemic (SPS Indonesia, 2020). In contrast, Suara Pembaruan has a lesser national daily circulation of approximately 172,000 with the majority (138,530) distributed within the Jakarta metropolitan area (Segara Advertising, 2020). Whilst both are national newspapers, their location in Jakarta is well suited for this study, where there was a focus on COVID-19.
The sample period of data collection followed the COVID-19 outbreak in Indonesia, when the Indonesian Government implemented large scale social restrictions [Pembatasan Sosial Berskala Besar, abbreviated to PSBB], the equivalent of a partial lockdown to restrict the spread of COVID-19 throughout the archipelago (Sutrisno, 2020). The Indonesian capital city region of Jakarta was the first province to implement social restrictions which were initially planned to be conducted in three stages (Suara Pembaruan, 2020). Part of the period of the three-phased partial lockdown was included in the sampling period of this study. We selected a four week period between 15 May and 11 June 2020, that captured part of the lockdown period of Phase II, all of Phase III and 7-days following the end of social restrictions originally scheduled for 4 June 2020. During this time, COVID-19 was particularly prominent in news reporting. As such, it is a suitable period to examine the use of language in response to the pandemic.
Sample data was accessed online but drawn from the print version format of each newspaper, as sold in hard copy. Our sampling technique involved collecting data from the front pages of the print version of each newspaper. Over the 28-day sampling period, there were 24 editions of Kompas and 21 editions of Suara Pembaruan. The discrepancy in total number of editions between the two newspapers is because the newspapers did not print on three public holidays during the sample period. Also, whilst Kompas publishes seven days a week, Suara Pembaruan is not printed on Sundays. These differences in weekly printing schedules resulted in different total editions sampled between the newspapers.
In analysing the sample data of this study, we apply a mixed method approach to examine quantitative and qualitative dimensions of the CS phenomenon. To examine the quantitative occurrence of pandemic related terms originating from English, content analysis was undertaken. This allowed us to identify key English lexical items, and to highlight occurrences of English CS. Content analysis is 'used to determine the presence of certain words within texts or sets of texts' to 'quantify and analyze the presence, meanings and relationships of such words' (Colorado State University, 2020). It is a recognised approach to quantify phenomena occurring in texts (Stokes 2003, p. 56) and a method that 'seeks to be systematic, objective, and transparent' (Drisko & Maschi, 2015, p. 3). We use this approach to present quantitative descriptions of the occurrence of English lexical items in the sample data. This allows scope to compare the range of alternative terms and their relative frequency of use. Beyond this quantitative analysis, subsequent data analysis applies a qualitative approach to explore specific examples in greater detail. This includes semiotic dimensions of the range of terms in use through CS, with a particular focus on the phenomenon of CS reiteration. Applying semiotic analysis to this phenomenon provides an avenue for in-depth qualitative analysis of language as signs or as possessing symbolic meaning. For the study of texts, semiotic analysis can be very useful when combined with content analysis (Stokes, 2003, p. 75). This mixed method approach offers breadth and depth of analysis -which is important in exploring the complex socio-linguistic dimensions of CS.

FINDINGS AND DISCUSSION The linguistic effects of COVID-19
The neologism 'COVID-19', an abbreviation of 'coronavirus disease 2019' was first coined on 11 February 2020 in a World Health Organization report, but by March 2020 'it was one of the most frequently used nouns in the English language' (Oxford Languages, 2020 p. 7). Consequently, the word COVID-19 quickly became a core part of many languages around the world, Indonesian included. In fact, in our corpus data the individual term 'COVID-19' represented 1.41% of all words in Suara Pembaruan, and 1.20% of total words in Kompas front page articles. Whilst a percentage of a little over one percent may not appear to be a large number, considering the fact that the term COVID-19 was only created two months prior to the data collection this percentage represents quite a substantial figure for a newly created word. This percentage may also be regarded as a broader reflection of the significant rise in usage of other COVID-19 related terms demonstrating the 'phenomenal breadth of language change and development during 2020' as a result of the pandemic (Oxford Languages, 2020, p. 3).
For this study, front page articles were grouped into a range of topics covering all aspects of life and society. A survey of topics illustrated that COVID-19 predominately dominated front page reporting, regardless of the topic, emphasising how the pandemic impacted all areas of life. Topics related to COVID-19 included, social, health, sociopolitical, socio-economic, education, politics, technology and sport. 81 of 92 (88%) of front page Kompas articles were related to COVID-19, while 52 of 55 (94.5%) of Suara Pembaruan reports were linked to the pandemic. A small percentage of non-COVID-19 related topics included Socio-political (Black Lives Matter), and socio-economic issues. 41% of KOMPAS articles were social COVID-19 related topics, while the largest percentage (36.6%) of Suara Pembaruan's reports were socio-political COVID-19 related articles (see Table 1). In the 24 editions of Kompas, a total of 28,415 words were identified in the front-page headlines and articles, with 303 (1.06%) classified as English. While in the 21 editions of Suara Pembaruan the total word count was almost 10,000 words less at 18,468, however the percentage of English present was double that of Kompas at 2.14% (397 words) (see Table 2). While there were hundreds of English words identified in both newspapers, this study is primarily focused on English pandemic specific terminology. Pandemic terminology was deemed a collective consisting of several words, for example 'stay at home', 'physical distancing', and 'new normal' etc. English specific pandemic related terminology was employed by Kompas on 21 occasions, while Suara Pembaruan had 99 instances. Although English pandemic related terminology was a small percentage of the total English in both newspapers, we chose to focus on how English pandemic specific terminology was employed, specifically in terms of CS reiteration. The reason for this is because CS forms 'one of the principal mechanisms of borrowing' which is the evolution of language change (Gardner-Chloros, 2009, p. 20 & 23).

Reiteration of 'New Normal'
The linguistic phenomenon of reiteration or translated paraphrasing represents a form of CS when words are 'reinforced, emphasized, or clarified where the message has already been transmitted in one code' (Eldridge, 1996, p. 306). Bhatia and Ritchie (2013, p. 384) explain CS reiteration as a message that is conveyed in one language and then 'is either repeated in the other language literally or with some modification to signify emphasis or clarification'. We highlight these modifications in the sample data from Kompas and Suara Pembaruan and the distinct lack of uniformity of Indonesian reiterations of individual English terms. The most striking example of code-switched reiteration translation in the sample data is the term 'new normal'. According to the Oxford English Dictionary (2020) the term new normal refers to 'a previously atypical or unfamiliar situation, behaviour, etc., which has become standard, usual, or expected'. The term was initially coined in a report post Global Financial Crisis of 2008 (International Monetary Fund, 2010, p. 12) but its recent use in relation to COVID-19 has been prolific.
In the sample data of this study, the term new normal was used by various groups ranging from journalists to politicians and was by far the most commonly used Indonesian code-switched reiterated English pandemic term on the front page of Suara Pembaruan with 38 occurrences, and second most in Kompas with five occurrences. Its frequent use is not surprising. It is interesting that from the 38 occurrences in Suara Pembaruan, new normal was translated 15 different ways, while Kompas had four different variations (see Table 3). As noted by Sneddon (2003, p. 192), 'some loan translations are quite long, revealing how prepared many journalists and others are to adapt English-language phrases rather than simply borrowing the English'. Such a variety of different translations has the potential to cause confusion amongst non-native English readers in understanding which term is actually the 'correct' translation. The Linguistics Encyclopedia (2002, p.

70) explains 'repetition' of words and phrases in
another language is a method used to 'clarify a point', while Foster and Welsh (p. 403, 2017) state reiteration within the Indonesian context 'seems to be for communicative effectiveness and emphasis'. The finding that the term new normal was frequently translated differently suggests it was a challenging concept to express.  (Leksono, 2020). Interestingly, in the same article, the author, the Chancellor of the Nusantara Multimedia University suggests if the English was correctly Indonesianised it should form 'kenormalan baru', the same translation proposed by Language Centre of the Ministry of Education and Culture (Leksono, 2020).
Following the Kompas article on new normal there were no further occurrences of the English term new normal, nor any further Indonesian translations apart from the Kompas editorially agreed term normal baru. Between 1 June and 11 June 2020, the Indonesian translation of normal baru occurred 36 times in the sample data from Kompas. While Kompas editors appeared not to adhere to the Badan Bahasa's recommended suggested translation of kenormalan baru, the editors did stand by their declaration in consistently using only the one translation. This was not the case with Suara Pembaruan, as illustrated in Table 3.
While the term new normal was associated with the most significantly varied translations, other English pandemic terms were also code-switch reiterated with different translations, albeit not as frequently. While the only discrepancy of varied translations identified in Kompas were of 'New Normal' as discussed above, the English term 'rapid test' was present three times in Suara Pembaruan but translated differently twice, while 'basic reproduction number', or R number, was identified 10 times in the same newspaper, and was translated into Indonesian differently on four occasions as detailed below (with Authors' English translations).
• One reason for the variation in translation could be that as a result of the pandemic, 'in a short period of time specialist epidemiological and medical vocabulary entered everyday discourse, such as the R number' (Oxford Languages, 2020, p. 5). While there were many occasions of CS reiteration which were employed as a method to clarify or explain the English, there were also numerous instances where English pandemic specific terminology was utilised with no reiteration in Indonesian. Those instances can be regarded as traditional CS. Very specific pandemic related English terminology such as 'lockdown', 'stay at home', 'Real Time Polymerase Chain Re-action', 'self-assessment', and 'exit strategy' (referring to loosening of restrictions) were used in both newspapers without any Indonesian translation at all.
Although CS is a common phenomenon in Indonesian newspapers, the use of newly introduced, and very specific English terminology related to the pandemic may cause confusion for many Indonesian readers, who are not proficient users of English. There appeared to be no consistency or uniformity by either newspaper in deciding which English pandemic related terminology would be followed by an Indonesian code-switched reiterated translation, and those which would not. Our data was collected relatively early during the pandemic, which could be a reason for the confusion and inconsistency in deciding how to translate foreign terms as the linguistic situation was still fluid and developing.
Conversely, there was many instances of what may be regarded as superfluous CS reiteration, where general English words were code-switch reiterated with an Indonesian translation that was seemingly conceptually or communicatively redundant.
Examples include 'scoring' [perhitungan] in Kompas on 19 May 2020, along with 'decline' [mundur] in Kompas on 11 June 2020. This CS reiteration may be regarded as redundant because meaning could have been adequately conveyed by the equivalent Indonesian term. Sneddon (2003, p. 179) has noted cases of this nature, where 'it is difficult to see if anything is achieved by this'. Hassall (2011) also notes this phenomenon where borrowed items do not fill a lexical gap but exist alongside an Indonesian word of the same meaning, thereby creating a synonym which can potentially change the language. With no apparent communicative need for this, we suggest such CS reiteration is for semiotic purposes, drawing on the perceived prestige of English to convey a sense of credibility or an authoritative voice. Signalling the source of information is international could be perceived to be particularly advantageous in the context of COVID-19. In addition to CS reiteration, there were also instances where the Indonesian translation of an English pandemic related term was used without any reference to the English, or mixed with English, therefore being employed as a borrowed term which had been Indonesianised through translation. An issue that arises with this phenomenon is with inconsistencies in the translation, as detailed above with the variations of New Normal. On 3 June 2020, the medical device known as a 'infrared thermometer gun' was translated differently in the two newspapers. Kompas used the term termometer tembak (lit. firing thermometer), while Suara Pembaruan used the term termometer gun. Whilst Kompas used only Indonesian to explain the foreign concept, Suara Pembaruan code-switched with English employing the word 'gun'. Neither newspaper was in line with the Badan Bahasa (Badan Pengembangan dan Pembinaan Bahasa, 2020b), who suggests the precise Indonesian translation of the term is pistol termometer [thermometer pistol]. This discrepancy in translation demonstrates Badan Bahasa's difficulties in countering the influence of English technical terms. One of the reasons for the failed adoption of Badan Bahasa's proposals is due to the nature of time sensitive news reporting in the mass media. Sneddon (2003, p. 150) explains journalists 'rarely have the time for elaborate language or for finding out what the terminology creators suggest for a particular word or English concept', which consequently creates a situation of core borrowing where English words are 'spontaneously' borrowed or adopted rather than through 'planned' or 'deliberate' borrowing. Sneddon (2003, p. 150) further explains that while respected 'newspapers like Kompas' 'are serious about the quality of their [Indonesian] language' use, they rarely strictly abide by the terminology guides from Badan Bahasa and often 'tend to create their own terms and make stylistic innovations when they feel the need and many of their creations enter general usage, irrespective of the wishes of language planners'. Evidence of this can be seen with Kompas' effort in employing the term termometer tembak. Whilst technically employing only Indonesian words, rather than CS with English like Suara Pembaruan, the term termometer tembak is not the standardised form recommended by Badan Bahasa.  Kompas and Suara Pembaruan, it is the Indonesian public who inevitably 'adopt or reject a neologism' through an 'implicit democratic process' (Hardini & Grange, 2016, p. 158). It is the Indonesian public in the broadest sense that constitute the speech community of the Indonesian language. Of course, language may be influenced by politicians, government and the mass media, along with an almost infinite range of other entities. As a natural process, language ultimately evolves at the behest of its speech community as a collective. It may overrule a government's will to linguistically regulate. This democratic process is reflected in the print media's varied vocabulary choices, be they English, Indonesian or neologisms derived from other languages. This is particularly true now since the Reformasi Era, in contrast to the New Order era when the 'government saw code-switching as a form of disorderliness', symbolising a position that was 'un-Indonesian' (Martin-Anatias, 2018, p. 341).
Despite the Badan Bahasa's efforts, English terminology is still often adopted by the speaking public. This study illustrates the role the mass media plays in this process. The adoption of borrowed words depends a lot on the 'openness of the receiving speaking community' and this process is often 'difficult to define, as it involves history, international relations, collective opinion and clichés about foreigners, sociology and governance' (Hardini & Grange, 2016, p. 158). The Indonesian language is dynamic, and reflects historical influences of Arabic, Dutch, Portuguese, Chinese, Sanskrit, English and local indigenous languagesmost obviously Javanese. It is this eclectic mix of lexical influences that largely sets it apart from other Malay dialects, where the very essence of the Indonesian language is 'characterized by its absorption of loan words and readiness to create neologisms' (Amerl, 2006). Universitas Indonesia lecturer in communication, Aimee Dawis notes that 'the dilution of Bahasa Indonesia is not the result of a deliberate government policy, it's just occurring naturally' (Onishi, 2010).
The Badan Bahasa recognises that foreign languages, in particular English, have 'permeated various aspects of the [Indonesian] people's and nation's lives' (Murtisari & Mali, 2017, p. 92). As a government apparatus, the Badan Bahasa has had the role of trying to 'manage' the development of Indonesian languagewhich is a very difficult task. Given the natural evolutionary nature of language itself, overseeing language development is an ongoing process. It perhaps should be more a task of keeping up with language change, and trying to guide the evolutionary process rather than control it. While policy can certainly enable or accelerate language change processes, language policy alone may have little effect. If policy is in line with natural processes it can be far more effective as a developmental tool, however natural language change often happens in unpredictable ways, which is why language planning is not always successful. As detailed in this study, the influence and power of mass media plays a significant role in the borrowing of English language into Indonesian and as a consequence intrinsically changes the Indonesian language.

Code-switching reiteration: A vehicle for language change
Put simply, CS involves borrowing lexical items. The use of CS assumes that the borrowed term will be comprehensible to the intended audience. Yet, for broader audiences such as readership of a newspaper, it is impossible for the writer to know the comprehension abilities of all readers which will be many and varied. CS reiteration is a way to address the incomprehensibility of borrowed words, by providing a translation. CS reiteration can therefore be regarded as an effective communication technique.
The next evolutionary stage of language change is when a loanword becomes more integrated into a language through adaptation, as observable for phonological reasons. 'All languages borrow words or phrases from others with which they come in contact' and 'codeswitching may often be the first step in this process' where words then become more widely used and accepted into the recipient language (The Linguistics Encyclopedia, 2002, p. 69). CS involving foreign terms is acknowledged as not only enacting language change in lexicon (Gardner-Chloros, 2009;Mahootian, 2005) but also as having the potential to generate structural or syntactical change (Backus, 2005).
This study has also discovered a range of different words in Kompas that were not technically English, since they had been morphologically adapted, but also could not be identified in the official Indonesian language dictionary [KBBI]. Words such as militerisasi, memiliterasi, liminitas, and skrining were all found in Kompas, and were used as standalone words with no English CS reiteration. This usage suggests they are borrowed terms that have been altered or Indonesianised. The adaptation of foreign words into Indonesian involves the adopted word being altered through the affixation of morphemes as prefixes and/or suffixes.
These are examples of how Indonesian journalists increasingly employ 'English words in their writing, including English root verbs that are adopted and modified with the addition of Indonesian verbal affixes' (Foster & Welsh, 2017, p. 400). Sneddon (2003, p. 191) notes 'heavy borrowing from English has resulted in a number of other innovations' in word formation where English words 'form Indonesian verb bases'. Examples within our corpus were the words militerisasi, and memiliterasi. Neither appear in English or Indonesian dictionaries. The word militerisasi, presumably from the English noun 'militarization' has been Indonesianised into a verb through verbal affixation of the prefix me-, to mean 'to militarise'. Sneddon (2003, p. 191) adds, 'there is now a large number of verbs whose base comes from a derived noun (in English)'. The example of memiliterisasi illustrates the 'dexterity of the Indonesian language in adapting loan words from another language, and how this dexterity can be creatively applied by journalists' (Foster & Welsh, 2017, p. 401). This influential role of journalists is noted by Hassall et al. (2008, p. 58) who claim that most English loanwords that enter the Indonesian lexicon 'are first coined by journalists or by other educated language users who are familiar with English'. The identification and use of borrowed English words in Indonesian, as this study has documented, highlights the critical role that the print media plays in language change.

CONCLUSION
This study has demonstrated how English pandemic terminology has been employed by two prominent Indonesian newspapers -Kompas and Suara Pembaruan. Quantitative and qualitative analysis has shown how the Indonesian lexicon has been influenced by English as a result of the COVID-19 pandemic. There is clear evidence that the Indonesian language has been influenced by English pandemic vocabulary. This raises questions over the potential longer-term effects of this period where numerous new lexical terms were rapidly introduced into the Indonesian language, largely through English code-switching.
The evolution of languages is a natural and ongoing phenomenon. When languages come into contact, it is inevitable that elements of one language will be switched, borrowed or absorbed by the other language. This process applies to any language and it is highly likely that the COVID-19 pandemic has enabled similar linguistic influences of English upon other languages.
This study focuses on the case of Indonesian and demonstrates how COVID-19 has expedited the borrowing of English pandemic-related terms. We argue that this process generates increased opportunities for language change, and that the Indonesian media is a key driver of this process, through code-switching. Due to the circumstances of the COVID-19 pandemic, and increasingly globalized actual and virtual worlds, the prolific use of English by Indonesia's mass media has contributed to a rapidly increased influence of English on the Indonesian language. This influence is not new, but the phenomenon has grown over the past 20 years since democratisation and significant political reform in Indonesia [Reformasi]. In 2020, COVID-19 has rapidly brought into focus the use of English terms through code-switching in Indonesian. This lexical influence is likely to have an enduring effect and enact change to the Indonesian language, yet such a longer-term assessment will have to be judged by future linguistic research. What is clear at this point, is that COVID-19 has led to a rapidly accelerated lexical influence of English on the Indonesian language, largely enacted by the Indonesian media through code-switching. It is likely that this process will produce enduring lexical change and contribute to the broader issue of language change, that potentially extends to matters of syntax and grammar. In managing the development of the Indonesian language, the Badan Bahasa has a challenging task to carefully consider 'democratic' language choices of its speech community that generate organic yet potentially 'chaotic' language change, rather than a more traditional prescriptive approach to language development.