“Such a Good Night”: Analyses of Korean-English code-switching and music video comments of ASTRO’S songs

To attract global multilingual audiences, English lyrics had been included in K-Pop songs for decades as a strategy (Lauren, 2018; Sea, 2018). This paper analyzed the English usage of “All Night”, “Blue Flame”, and “Knock” by ASTRO, a third-generation K-Pop group from a non-mainstream agency to examine code-switching and the global recognition of K-Pop songs. All songs were examined by using code-switching theories from Poplack (1980) and Bullock and Toribio (2009) and other theories on function analysis by Lawrence (2010) and Lee (2004) and comment analysis by Fernandez-Martinez et al. (2014) and Kwon (2019). The findings revealed that the majority of code switches in “All Night” and “Blue Flame” were inter-sentential, yet “All Night” and “Knock” were more similar in utilizing English by having a new function (i.e., emphasizing a dialog from one’s perspective), an important addition to Lee’s (2004) functions. Furthermore, from the usage and function analysis and the comment analysis, the audience responses seemed to be influenced by the singers’ mispronunciations and the non-existence of a new function that failed to create uniqueness in English usage in “Blue Flame,” which may have led to the song’s failure to win awards. In conclusion, the research suggested that, in the near-absence of promotional support from the agency, the interplay among the number of switches, accurate pronunciations, and a new English-switching function seemed to play an important role in making “All Night” and “Knock” popular among multilingual audiences.


INTRODUCTION
The inclusion of English words and phrases in K-Pop lyrics has been a phenomenon for decades. Since the 1990s, English has been utilized as K-Pop song titles, song lyrics, and singers' stage names. It is used to increase the singers' reputation and to approach international audiences, and it seems that it has helped the spread of the Korean wave worldwide (Lawrence, 2019;Willis, 2014). Nevertheless, according to Willis (2014), although it has been practiced for a long time, inserting English in K-Pop songs is still seen as problematic in the eyes of the about the singers' physical album sales but also the number of global votes, followers, and music streaming (Kong, 2016). Thus, English usage seems to increase votes and followers, in the end, the probability of winning awards. In winning awards, K-Pop fans revealed that the awards could be achieved with the agency's support and promotional strategies (Koreaboo, 2016). Singers or bands from big agencies were more likely to achieve more awards, but not much for singers who debuted by indie or small agencies. Then, it is the focus of this research to explore the inclusion of English in K-Pop songs by singers supported by small agencies in globally grabbing multilingual audiences' attention.
This English inclusion in Korean songs projects a behavior of language shifting in different domains by a particular participant that Bullock and Toribio (2009) and Holmes (2013) called code-switching. According to Poplack (1980), there were three types of code-switching. First, intra-sentential switching was a shift from one code into another within one phrase, clause, or sentence. Both languages' grammatical structures overlapped. Second, for intersentential switching, the alteration of one code occurred between utterances or sentences that were separated by a full stop. In other words, it constructed an individual sentence for each code. Intra-sentential switches were uttered by bilinguals who were advanced in both languages (Holmes, 2013). The last type of code-switching was tag-switching. It was commonly produced by bilinguals whose skills were limited in one of the languages, the same with intersentential switchings (Holmes, 2013). It typically occurred as an additional expression that contained one's cultural words, including idioms, fillers, exclamations, and interjections. In addition, tagswitching was considered flexible, meaning the word or phrase could be inserted in any position without disarranging the sentence (Bullock & Toribio, 2009;Poplack, 1980). The following figure shows how different types of code-switching used in Korean song.

Figure 1
Examples of Different Types of Code-Switching in Korean Songs (Sea, 2018, pp. 26-32) Intra-Sentential morachin pokpung soge nae maeumeun Yeah I'll be just fine Inter-Sentential tto neol gidarigetjiman (I'll wait for you again) boy I can't cover up my heart neol bomyeon nan useumi saeeo nawa (when I see you, laughter seeps out)

Tag-Switching Oh ije nan eodiro
Oh now where am I?
In previous studies, K-Pop songs had been analyzed for their English utilization. A study by Sea (2018) utilized the code-switching theory by Poplack (1980) and determined the motivation behind codeswitching on songs of an album My Voice by Taeyeon, a singer from a reputable agency called SM Entertainment. Another study by Jocelin and Tryana (2019) applied code-switching theory in the song "Taste" by a girl group from the same agency called Red Velvet. The research classified each type of code-switching and code-mixing, then reconstructed how English was used in the song. The theory of code-switching is also applicable outside the Korean context, such as a study by Rusli et al. (2018) that used the theory to detect and uncover functions or motivation behind the usage of English lyrics in Malaysian songs, similar to research by Likhitphongsathorn and Sappapan (2013) in Thai songs and Kadir (2021) in Indonesian songs.
In addition, some studies conducted research by determining functions of English mixing based on Lee (2004), who stated that English was implemented in K-Pop songs as stylistic devices, attention-getters, loanwords, assertion of sensuality, assertion of unsettled identities, and assertion of resistance. A study by Lawrence (2010) analyzed twenty-four randomly selected songs of the second K-Pop generation singers. The research identified English variations, linguistic structures, and English usage functions based on Lee (2004), namely to assert selfidentity, create new meanings, challenge dominant representations of authority, resist mainstream norms and values, and reject older generations' conservatism. It also utilized the function to identify whether the English was standard (e.g., American or British), Konglish, or some other variety of English. Likewise, the results of studies by Dovchin (2017) and Lauren (2018) were in line with Lee's (2014) findings, which were mostly for stylistic purposes, asserting sensuality, and asserting resistance.
The aforementioned studies only examined the implementation, function, and motivation of inserting English in K-Pop songs, mostly from a big agency. Also, they did not do further research on the impact of inserting English in the song as a strategy to gain popularity. Thereby, this study aims to fill a gap by considering the global multilingual audience's response. The result of the study contributed to showing whether English insertion in a K-Pop song appealed to the audience or not.
Adapting previous studies (Jocelin & Tryana, 2019;Likhitphongsathorn & Sappapan, 2013;Rusli et al., 2018;Sea, 2018), Poplack's code-switching theory (1980) was employed to find out the placement of the switch and the types of code switches, as well as identify linguistic structure, including phonetics, phonology, morphology, and syntax. The phonetics and phonology were discussed mostly in the comment analysis by comparing the standard English with mispronunciations heard by the audience, whilst all morphology and syntax discussions were in the code-switching analysis and English usage and function sections to examine how proper the English used in the songs and whether there are Konglish or not.
Furthermore, Lee's (2004) types of English mixing were chosen as the tool to indicate functions. Prior studies (e.g., Lauren, 2018;Lawrence, 2010) used Lee's (2004) functions for songs from the second K-Pop generation group, so ASTRO were selected as the representative of the third K-Pop generation (Song, 2020). The group was chosen as the subject of this research due to their contract under Fantagio, which was a small entertainment agency. This meant that ASTRO may have to rely more heavily on English lyrics in their songs to obtain global recognition due to the limited resources Fantagio has to market the group. In order to complement that, this research clustered the YouTube comments by their sentiment towards the English lyrics. The provided comment section on YouTube assisted the creator to obtain the audience feedback data (Siersdorfer et al., 2010). Since there are hardly any comment analysis papers on codeswitching, we looked at studies reporting on audience comments on other kinds of corpora. A study by Fernández-Martinez et al. (2014), in which they studied audience perception towards car commercials, grouped comments according to whether they were about the audio, visual, or other features of the commercials. Another study by Kwon (2019) looked at comments by multilingual audiences on North Korean musical performances broadcast on YouTube. Viewers from a multitude of countries, including Russia, Germany, South and North Korea, Italy, Spain, France, Portugal, Japan, China, Vietnam, and Thailand, commented on various musical performances by North Korean children and the comments were classified as positive and negative. Hence, this research answers the following questions: (1) What types of codeswitching occur in "All Night", "Blue Flame", and "Knock"? (2) What were the differences in terms of English usage function between each song? (3) How impactful was the English usage in each song in engaging global audiences?

METHOD
A qualitative research design was used to identify the used code-switching in ASTRO's lyrics, the English usage function in the lyrics, and the engagement from audiences in the comment section of the video clip. Three ASTRO's songs entitled "All Night", "Blue Flame", and "Knock" were selected in this study. The reasons for the selections were that "All Night" was the first song that gave them their first victory on a music show, The Show. After that, ASTRO released "Blue Flame" but did not win any awards although many people opined that this song was one of their favorites as the predominant comments on an Allkpop forum showed that "Blue Flame" was the song of the year (Vyuna, 2019). Moreover, another forum discussed the song in which the overall comments stated that it was a nice and memorable song (Minghao_life, 2020). Then, ASTRO released "Knock" and won its second award at Show Champion. Thus, these research data of English mixing songs were chosen to compare each English usage and function to note the possible factor behind ASTRO receiving a variable number of awards.
To answer the first and section research questions one and two, the lyrics from each song was parsed per line, and the types of code-switching were determined according to Poplack's (1980) codeswitching classification (intra-sentential, intersentential, and tag-switching). After that, English mixing functions were analyzed based on Lee's (2004) categories (attention getter, loan-word, assertion of sensuality, assertion of unsettled identity, assertion of resistance, and new function). The results of the analysis then inserted into tables for further discussion. As for the third research questions, the audience responses in the comment section of the video clip, the audience comments, mainly comments in English, were collected from each song's YouTube page from the official channel. According to Bland (2019), YouTube was the most used platform by K-Pop agencies and had become the base for K-Pop bands to attain worldwide popularity, provide complimentary access to any music content, and allow the audience to engage with the singers.
The first step of this research qualitatively identified the types of code-switching and their placement organized and presented by using a table. The types of code-switching were classified in accordance with Poplack's (1980) theory. For instance, overlapping codes in a sentence "I'm 베고파" was considered as intra-sentential switching, meanwhile two codes in different sentences but uttered by one speaker, such as "나는 배고파. Let's eat." was deemed as intra-sentential switching. Whilst, if a sentence of one code contained an exclamation of a different code, such as "Woah 나는 배고파. 먹차." was found, it was identified as tag switching. The identification of placement was based on song structure, such as an intro, verse, chorus, bridge, and outro.The second step examined English usage and function with a qualitative method, following Jocelin and Tryana (2019). Findings from the previous step were discussed for their morphological and syntactical aspects following Lawrence's (2010) study, and a study by Lawrence (2019) was employed as a guideline to delve into the occurrence of Konglish linguistic constraints. The part was done to depict whether the songs phonologically and grammatically implemented Konglish or standard English.
For the sociolinguistic approach, Lee's types of English mixing (2004) were utilized to search the function of the English lyrics so that the uniqueness of the three songs could be revealed. In addition, to examine the coherence of meaning in the lyrics as a whole, this theory involved the semantic level. As previously mentioned, some studies implemented code-switching to break down the English structure and elements. For example, in the sentence "I'm 배고파," the insertion of the adjective '배고파' or 'hungry' could be an example. In Korean, the subject and predicate "I am" is "나는" and if it is wholly in Korean, it will be "나는 배고파." Therefore, even if it is "I'm 배고파" or changed as "나는 hungry", the sentence is still grammatical although it is intervened by another language.
The final step of this research concerned the audience's response through comment analysis and the disparity between positive and negative feedback. The classification was done based on some keywords found in the comments. For instance, the keywords for positive comments included talented, elegant, and heavenly, for example, and for the negative ones, commenters wrote words and phrases such as kinda cruel, look sleep deprived, and made to practice very hard. Thus, adapting the previously mentioned studies (Fernández-Martínez et al., 2014;Kwon, 2019;Siersdorfer et al., 2010), the clustering of each song's comments in a form of a table was divided into positive and negative sentiments, which was manually counted and depicted in numbers and percentages. Subsequently, all findings were discussed to relate them to the results from the previous steps, including the phonetics and phonology aspects by comparing the standard English with mispronunciations heard by the audience.

FINDINGS AND DISCUSSION Code-Switching used in ASTRO's songs
The most frequent code-switching in three ASTRO's songs are depicted in the Table 1. As can be seen in the table, the occurrence of intra-sentential are the highest code-switching in three ASTRO's songs (34), followed by tag-switching (21) and inter-sentential (15).

Song
Intra-sentential Inter-sentential Tag The most frequent code-switching type in "All Night" was intra-sentential switching, then followed by inter-sentential switching and tag-switching. Based on the song structure, shown in Appendix A, all switches could be found in the first and second verses, pre-chorus, bridge, chorus, and outro. There were seven intra-sentential switchings, which were in two forms of the grammatical unit: word form and phrases. Notwithstanding, the use of phrases in number 7 was slightly different from the rest in which a repetition of the phrase "pick it up" was detected. The similarity in both forms was they consistently code-switched at the end of a sentence. The second most frequent type of code-switching was intersentential switching. In the total of four switches, there were two variations of style. The first one was in the style of a repeated phrase, such as the lyrics "I can't wait I can't wait I can't wait" (numbers 5 and 10). Likewise, the repetition was available in number 13. The other distinctive style was single phrases "such a good night" (11) and "one last time" (12). For tag-switching, there were only two switches apparent. The word "oh" (3) was deemed as an interjection, whereas the repeated phrase "live it up" (6) was an idiomatic expression, expressing the feeling of excitement.
Details of the findings of code-switching in the second song, "Blue Flame" can be observed in Appendix B. Similar to the previous song, there were seventeen English codes to switch intra-sententially in the form of words and phrases. The majority of the word-level switches (numbers 2, 4, 23, 25, and 26) were at the last word of the line. In a different way, the word "cause" (5 and 15) was utilized as a conjunction of a separated sentence that was continued in the upcoming pre-chorus. Whilst in number 27, it was an initial word of a dependent clause. Yet for number 23, the verb switch "knock" was produced in the style of repetition. Switches in the form of phrases were present as well. Unlike the phrase "burn me" (3), some phrases were constantly mentioned throughout the song. Instances of these were "with your blue flame" in pre-choruses (6, 16, and 28); and "like a blue flame" in choruses (10, 11, 20, 21, 32, and 33). Then for inter-sentential switching, there were five switches detected. The majority of the switches were in the form of a sentence; however, the placement of them was various. The referred parts were at the beginning of choruses (7, 17, and 29); in the middle of the fifth verse (24); and in the closing of the last chorus (34). Another line that switched inter-sententially was "I yah yah yah yah yah" (1, 13, 14, and 22). Such a line repeatedly appeared in the intro, the hook, and adlibs. Lastly, there were eight inserted tag-switching whose role was as an interjection and fillers. The only detected interjection was the word "yah" (12). Meanwhile, the fillers could be identified through the line "running running" (8, 18, and 30); and "falling falling" (8, 19, and 31).
Finally, details of code-switching in "Knock" are depicted in Appendix C. In contrast to the others, this song had tag-switching as the most frequent type of code-switching and inter-sentential switching as the least one. The tag-switchings were categorized into fillers and interjections. In varied forms, the lyrics "fly up" (4) and "knock knock knock" (6, 8, 12, 14, 17, and 19) were deemed as fillers. For the other category, interjections could be seen through the switch of "yeah" (16) and "no oh oh" (15, 20, 21, and 23). Next, the established intra-sentential switchings, unit and style-wise, were as varied as other songs. The placement of three switches was in the hindmost part of a sentence, such as "up up up" (7, 13, and 18); "star" (2 and 9); and "starry night starry night starry night" (3 and 10). Dissimilar to the other, the switch "wake me up wake me up" (5 and 11) was placed at the start of the line. For the last type of codeswitching, there were only two instances of intersentential switching in this song and both were used in two opposite parts. The line "oh yeah hmm" (1) was in the intro of the song; on the contrary, the ad libs "oh night" (22) was in the outro. As for the remaining switches, all of them were presented in the first and second verses, pre-choruses, choruses and bridge.
The result through this analysis shows the songs consist of all three Poplack's (1980) code-switching types; however, each of them has different quantities for each type. "All Night" and "Blue Flame" shared similarities by utilizing intra-sentential switching the most, inter-sentential on the second, and tagswitching the least. The song "Knock", on the other hand, has tag-switching, followed by intra-sentential and inter-sentential. Based on Daoh (2016) (as cited in Rusli et al., 2018), the most frequent codeswitching type is inter-sentential switching, but the three songs show otherwise with none containing it as the highest number. The only aspect that is similar with previous studies, such as research by Jocelin and Tryana (2019), is how tag-switching is the least used type. Moreover, according to the song structure, the majority of the songs switched the code in the verse, pre-chorus, chorus, and bridge, similar to Lawrence (2010) findings. Nevertheless, certain songs used code-switching in a variety of sections. Some are placed in the intro of "Blue Flame" and "Knock", the outro of "All Night" and "Knock", and the hook of "Blue Flame". Therefore, the result of this study agrees with Lauren (2018) and Sea's (2018) findings, which shows the consistency of utilizing codeswitching as one of the strategies to attract global audiences for decades.

English mixing strategies
The strategy of English inclusion on the three songs was examined through the structural and sociolinguistic approach. The discussion of the major levels of linguistics contained the phoneticphonology and morpho-syntactic levels. Regarding the semantic level, it was included whilst searching on each function through Lee's (2004) six types of English mixing findings. The stylistic purpose typically occurred on the lexical level, which formed a rhythm of repeated words of one language or with the equivalent feature of the other language. For the attention-getters, they contributed to the semantic level analysis, so the inserted word showed whether it was justified or unjustified with the whole message of the song. Moreover, an English loanword occured at the morpho-syntactic level, and thus the findings assisted in showing which English words were glocalized in South Korea. Lastly, the assertion of sensuality, unsettled identity, and resistance occurred at the discourse level, which depicted the expressed meaning in the song. From those types, each song turned out to give certain numbers of finding, which is depicted in Table 2.
Before getting into the lyrics, the title of each song was discussed first because the pattern of these three song titles contained English. According to Lawrence (2018), there were three patterns of K-Pop song titles, which were Korean titles with English in parenthesis, English titles with Korean in parenthesis, and a combination of English and Korean titles without any parenthesis.
In this case, ASTRO promoted "Blue Flame" with the only one in English, whereas the others were English titles with Korean in parenthesis, which were "All Night (전화해)" and "Knock (널 찾아가)". The English titles were based on the main highlight of the song lyrics, whereas the Korean titles were chosen based on the main activity that was introduced in the songs, such as '전화해' meaning 'call me' and '널 찾아가' meaning 'I am going to find you'. The use of two languages in one song title was to balance the target market-inside and outside Korea-which gave advantages to their popularity and income (Lawrence, 2018).
Getting into the lyrics, "All Night" has a significant pronunciation difference from the writing, which is from "cellular phone" into "cell[ə] phone" or /ˈselə foʊn/. Compared to the American English pronunciation /ˈsel foʊn/, as the closest one, that additional sound of /ə/ had an influence on the Konglish use. Based on Lawrence (2019), Konglish tended to be made due to the language play by Koreans and became their new vocabulary. In this case, the new pronunciation could be caused by a lexical gap when a new term, mostly related to technology, was initially introduced. For example, they pronounced a handphone as handeupon (p. 90). By means of that, the /ˈselə foʊn/ or celleuphone could be deemed as having a similar case, as the additional /ə/ sound was inserted to match the syllable of the Korean writing system or Hangeul (p. 86). In addition, the shortened term from "Cellular phone" into celleuphone could be considered as the result of Konglish from morphological localization of clipping in which combining two words into one (Mihalicek & Wilson, 2011, as cited from Lawrence, 2019). Moreover, in "Blue Fame", possibly, there was a mispronunciation for the switch "running running" in the first mentioned line of the first chorus. The sound /n/ in 'running' was pronounced more likely as /m/, so it was heard as /ˈrʌm.ɪŋ/. Although the word 'running' already existed in Korean system writing or Hangeul as 런닝 or reoning, this case leaned towards the phonological localization, similar to "All Night". Based on Widyadewi et al. (2021), the Korean nasal sounds consist of /n/, /m/, and /ŋ/. Therefore, it is possible for the three sounds occasionally to be heard as one another since they are produced from the same place.
In accordance with Lee's (2004) types of English mixing, all songs were in line with stylistic purposes. Eight lyrics to be found having this function in "All Night". Some examples could be depicted through the result of rhyming the last syllable of each sentence with other sentences before or after it. In the following instances, the sound that got rhymed was [i] or [ɪ]. The last syllable of "시간이" or 'sigani' (a) was rhymed with the word "sleep" (b), as well as "테니" or 'teni' (d) with the last vowel of "sending" in the next line. Additionally, at the syntax level, the phrase "don't sleep" (b) could be interpreted differently, perhaps as an exclamation. If the English translation lyrics were inserted, it altered the phrase into a sentence due to the added subject 'I' before it. Another example could be seen in the phrase "Cellular phone" (c). The word 'phone' rhymed with the last Korean word "척" or 'jeok' in the subsequent line. Despite the difference in the writing, both were pronounced as [ Similarly, "Blue Flame" also used rhyming for stylistic purposes in which there were five lyrics identified. The rhythmic style was formed by the English switches (g). Each verb and suffix "-ing" were repeatedly presented within and betweensentences.
g Another example of rhyming covered four lines of lyrics at once (h). The repetition of the sound [o] was visible in every repeated word at the end of the lines, such as "또 또 또 또" or 'ddo ddo ddo doo', "knock knock knock knock", "널 널 널" or 'neol neol neol', and "더 더 더 더" or 'deo deo deo deo'. At the syntactic level, the line "knock knock knock knock" portrayed a resemblance with the case in "All Night". The meaning of the line could be interpreted as an onomatopoeia of a knocking door for non-Koreanspeakers. The word 'knock' comes to be a verb form owing to the additional subject 'I'.
h. 멀어져 가네 또 또 또 또 Whilst in "Knock", the four lyrics below could be deemed as stylistic purposes as well because of the consecutively mentioned words and phrases, such as 'starry night' (i), 'wake me up ', 'knock', and 'up' (j In addition, there was a significant comparison between the word 'knock' in this song and in "Blue Flame". The word 'knock' here phonologically imitated the sound-or onomatopoeia-of a knocking, which was supported by the non-existing subject before it. The second paralleled function is attentiongetters. There are only two lyrics found in "All Night". The majority of the English switches were semantically connected with the rest of the lyrics, except for the phrase "one last time" in the chorus. It could not be justified because it was solely uttered as a cue for the song to be played one more round for the outro. Whereas, the instance of the justified lyrics could be seen in the line "such a good night" by the virtue of the meaning continuity from the previous lyrics. Moreover, the lyric was uttered within a specific gap of silence in between the instrumentals. k. 어느새 고운 숨소리 속삭이듯 잠이 드는 너 (Then I hear your breathing, you fell asleep with the sound of a whisper) Such a good night (ASTRO, 2019; ASTRO 아스트로, 2019) Apparently, certain aspects of English utilization in "Blue Flame" were the same. One line of lyrics "I yah yah yah ya yah" was deemed as an attention-getter but unjustified with the whole message. This line was constantly uttered, especially in the opening, the hook, and as the ad libs. It was due to its intention as a hook of the song. A hook was applied to engage the audience's attention (Lawrence, 2010).
l. 이제걱정 No oh oh (Don't you worry no oh oh) 어둔 밤은 끝이 날 거야 우릴 스쳐 지나갈 거야 모두 새로워질 거야 No oh oh (The dark night would end, it will leave us, we will be brand new no oh oh) (Oh Night) (ASTRO, 2020; ASTRO 아스트로, 2020) As for "Knock", there were three lyrics that was identified as attention-getters. In the switches of "no oh oh" (l), the lyric was constantly uttered four times throughout the songs and varied results of the semantic analysis were noted. The first "no oh oh" was semantically justified because the word 'no' acts as an affirmation of the prior Korean phrase of "이제 걱정" or 'don't you worry'. In opposition, the second "no oh oh" cannot be justified due to the unrelated meaning with the rest of the Korean sentence. However, the sentence was semantically more fitted with the line "Oh Night" below it because of the containing theme. The technique of the utterance "Oh Night" was similar to what had been done in "All Night". The placement of the line was in a faded out beat between the instrumentals.
At the discourse level, the third agreement with Lee's (2004) types is assertions of unsettled identity, which illustrate a dualism. Referring back to the lyrics in (h), the Korean lyrics depicted the person as a hopeless romantic. The first three lines explained how the person would pursue the significant one although he got ignored. Nonetheless, a sudden shift happened in the next two English lines: "don't blame it on me" and "you brought the flame in my heart." It indicated a passive-aggressive behavior by the uttered defensive sentence to not blame him for maintaining feelings. Subsequently, the next Korean lyrics expressed that the love would keep getting deeper in the long run.
Other than all the findings above, a newly discovered function was discovered: the use of English switching to emphasize a dialog from one's perspective. This was supported by Holmes (2013), stating bilinguals tended to switch codes within an utterance to discuss a certain topic, including quoting a person's speech event. Such a function was found in "All Night". In the last line of a lyric fragment (e), it contextually explained the situation of both interlocutors who were just starting a phone call. The greeting of "good morning" was expressed by the singer after picking up the phone. Another example was located in lyrics (k). The two lines of lyrics above were paralleled message-wise. The phrase "such a good night" referred to the whisper of the other interlocutor over the phone. What is more, ostensibly, "Knock" was also having the same new function. Noting the theme and the lyrics (i), the repeated phrase "starry night starry night starry night" was initially said by the seeker; however, the two lines were more portraying a dialog from the lost one's perspective, predictably an imitation of one another. Thus, the two lines of lyrics were the lost person's perspective telling the audience how the seeker's repeated calling was being heard by imitating the calling. It was deemed as a calling due to the sentence structure: present progressive tense, which indicated a person's calling by the nickname 'starry night' rather than merely introducing a nickname. Thereby, the switches helped to emphasize the meaning of the word 'star' and the phrase 'starry night'. Because "starry night starry night starry night" was the lost person's line, it could be interpreted as both sides considered each other as a star, which was believed as a hope or a special person that would save one another All in all, each song had a diverse result. Most of the English switches were primarily in the form of verbs, nouns, and interjections. Also, "All Night" and "Blue Flame" had erroneous pronunciations, but there was a difference between both discoveries. The finding in "All Night" could be deemed as an English localization for still having a recognisable meaning despite the additional sound to match Korean syllable, whereas the error in "Blue Flame" was the contrary. By changing a consonant sound in a word, it made a meaningless word. In relation to the function, English switches in the three songs were mostly in line with Lee's (2014) stylistic purpose. They were depicted by rhyming and repeating some lyrics. Nonetheless, the function that benefited the songs is in the English usage for attention-getter because those switches were intended to be the highlight and to make the audience remember them. Moreover, a newly found function in "All Night" and "Knock" was apparent in the form of dialog. The use of English in this new function was used to imitate an interlocutor's dialogue. Furthermore, the English lyrics were used grammatically and pronounced well, and there was no Konglish found in morphological and syntactic levels.

Comment Analysis
The data for comment analysis were obtained from each song's YouTube comment section on ASTRO's channel. Owing to this paper's English codeswitching and element analysis that discussed English switches, usage and function of the lyrics were taken into account. In total, there were around one hundred and forty-five English comments as of November 20, 2020. They included twenty-two comments for "All Night" (Table 3), fifty-seven comments for "Blue Flame" (Table 4), and sixtyseven comments for "Knock" (Table 5). The comments were divided into the sentiment of positive and negative, as well as each total and percentage. The positive comments included praise and good impressions towards both English lyrics and skills, whereas the negative sentiment included some misheard lyrics and struggles from the audience.  Table 3 shows that there were twenty-one comments in total that talked about English linguistics, usage, and function. The number of positive comments towards the song was fifteen. In terms of the lyrics, the line "such a good night" was the most frequently mentioned lyric by the audience. Some of them said that the line was the best part of the song and even replaced ASTRO's memorable lines in their earlier songs. Furthermore, several audiences opined that the lyric "one last time" and repeated phrases "live it up live it up" and "pick it up pick it up" were enjoyable to be heard. For the linguistic aspect, there was one comment that praised the rapper in the second verse who had a good pronunciation and flow. Nevertheless, there were six comments that showed some audience had misheard one of the lyrics, which was the lyrics "live it up live it up" being heard as "level up level up". This showed that there is an opposition with the finding and the discussed lyrics celleuphone from the previous section. None of the audience seemed to contradict it because of the use of the Konglish usage that is accepted by the audience. Compared to the case of "level up level up", the background of the commentators whose mother tongue is not English may influence the reason why it was misheard since there were not any pronunciation errors found.  Table 4 illustrates the comments analysis for "Blue Flame." The constantly discussed lyrics were the repetitive lines such as "I yah yah yah yah ya", "falling falling", and "running running" in the song. Those lyrics were the most memorable, especially the line "I yah yah yah yah ya" that showed the accuracy of using a hook in the song. Moreover, the line that contained the phrase "blue flame", for example "light a blue flame", "with your blue flame", and "you're my blue flame", were considered as memorable because it was oftentimes mentioned in the song. The fragment of the rap part "don't blame it on me you brought the flame in my heart" was the last discussed line. Some audience members gave positive feedback towards the rapper, the same rapper as the lauded one previously, by virtue of his English pronunciation. There were some comments that praised the pronunciation of all personnel in singing the English lyrics, yet still some audience misheard one of the lines. The specific line was "running running", which becomes "ramen ramen", in the first chorus. This issue matched with the discovery in the English usage and function analysis. The sound of /n/ was also heard as /m/ by some audience, which showed the error comes from the singers themselves. contradicting what occurred in "All Night". Considering the comments that talked about the line had a bigger number, this might be due to the change of meaning entirely. From 'running' that represents an activity to 'ramen' that is classified as food, it may support the reason why it catches the audience's eyes. Other than that, some comments from international audiences conveyed their struggle in remembering the whole lyrics due to the limitation of the language, such as the audience who are non-Korean speakers, so they do not know the meaning nor how to pronounce the lyrics. They could only memorize the English lyrics, and one of them added the wish for ASTRO to produce a full English song. As displayed in Table 5, there were sixty-seven collected comments that talked about the English usage in this song. Likewise, the repeated words and lyric in the song were the most anticipated ones by the audience. The line "knock knock knock" was highly discussed by the audience, followed by "starry night starry night starry night", "wake me up wake me up", "up up up", and "no oh oh". Those were considered as the most memorable lines. For this song, there were comparatively misheard English lines, such as "wake me wake me up" becomes "bring me up bring me up". Another misheard lyric was "oh night", which comes to be heard as "all night". In addition, there was a comment that talked about the linguistic element. The comment was questioning why Korean singers tended to pronounce "starry" as "stairry', with a voiced [ɪ] vowel in between the syllables. In accordance with the prior study, Lawrence (2019) argued that it was factored by the inclusion of Konglish in the phonological level. English alphabets were oftentimes introduced since adolescent in Korean writing or Hangeul to explain the word production concisely. As an example, the letter 'A' was pronounced as /ei/, instead of /ae/ (p. 86). However, those misheard lyrics contradicted the finding in the previous section with no error found. Similar to the negative comments for "All Night", the misheard lyrics mostly came from the audience whose first language is not English, so it could not be counted as an error from the singers' side. For English usage, similarly, some audience found a great problem in remembering and interpreting the whole lyrics precisely due to the limitation of their language.
The final results showed that more positive comments were given by the audience. The majority of them mentioned that the repetition of English lyrics and the song highlights that were in English, made them easily recall those parts. This indicates that the use of English for stylistic purposes and attention-getters worked. By the application of this analysis, it could be seen that the audience also paid attention to how ASTRO used English in each song. Mixed feedback regarding pronunciation could be noticed. Some comments signified the audience's fascination with the singers' English pronunciation, even wished for ASTRO releasing an English song, yet some comments revealed several misheard lyrics by the audience. Some lyrics for all songs seemed to get misheard by the audience, but only the comments for "Blue Flame" matched with the finding in the previous step of analysis. Meaning, it was valid that there was a pronunciation error coming from the singers.

CONCLUSION
This research focused on examining the types of code-switching, investigating the differences in English usage and function in ASTRO's three songs, and looking at the global engagement through responses towards them. By mainly using the sociolinguistic approach in the context of world Englishes, with additional structural approach and comment analyses, the findings indicate that the quantity or the quality of English is more conducive to the success of the song, provide a possible new function of English in songs by third-generation K-Pop groups, and introduce a new method from prior studies to analyze audience response from YouTube comment section to K-Pop songs. The result showed that the three songs have similarities and differences in switching codes; some corresponding functions and a new function are discovered. Moreover, various responses from the audience were observed, which indicates the importance of practicing proper English in English mixing songs. The analysis also found that the three songs only consisted of very few Konglish phrases in their lyrics. The highest number of switches was found in "Blue Flame," followed by "Knock" and "All Night". Among the three, "Blue Flame" is the only song that does not receive any award from a music show after the global voting system is implemented. This suggested that the highest quantity of switches did not determine the success of a song in the eyes of the audience, and there might be other factors contribute to winning awards. The findings also imply that regardless of the agency's status and reputation, various English utilizations as a lingua franca in K-Pop songs to reinforce global recognition may have affected each song's popularity to a certain extent. In this case, ASTRO gained a variable number of awards for "All Night", "Blue Flame", and "Knock", which encouraged research into finding out if English lyrics insertion in each song is one of the factors causing this variability.
On the one hand, the songs "All Night" and "Knock" had their strengths in good English pronunciation, having a new English-switching function, and favorable comments from the global audiences. These strengths, in turn, could have contributed to their winning awards. On the other hand, "Blue Flame" had more problems with English pronunciation, lacked a new English-switching function, and had more less-favorable comments, which could have contributed to its failure to win awards. Furthermore, all of the findings in this research enable future researchers to do deeper research on the English-switching function, such as comparing Korean songs from different genres or artists, or comparing multiple theories on Korean songs and then, perhaps, finding more new functions. Another choice is analyzing further the impact of English-switching on Korean songs or other Koreanrelated entertainment, which can be more accurate with the help of popular metrics, for instance, by displaying the real-time analytical subscriber's growth of the artist's YouTube channel before and after the release of each song.
For ASTRO, one factor that could have driven the success of their songs could be the uniqueness of English usage and function and English pronunciation. Despite the findings that this thirdgeneration K-Pop group still utilizes English in similar ways with previous generations, a new function can be found in "All Night" and "Knock", which represent the development of English utilization in recent generations. The similarity between the English functions in the three analyzed songs and those of previous generations is the predominance of, in line with Lee's types of English mixing (2014), stylistic purpose by rhyming the vowel of the last word in several lines and repeating a word or two in one line and attention-getter because of the justified English lyrics semantically align with the whole message of the song and the unjustified ones are for fillers. Another parallel function is seen in the song "Blue Flame", which illustrates the assertion of unsettled identity. However, a new function can be found in "All Night" and "Knock", which illustrates a direct dialog of one interlocutor towards the other, including in the form of a phone conversation and a calling. The newly found function widens the probability for them to get awards, and presents an important addition to the functions proposed by Lee (2004). For audience responses, considering the outcome of the audience's sentiments towards English elements, all songs receive more positive comments than negative comments. The most common positive feedback was in the matter of how remarkable some English switches were, especially the ones in rhymes. This proves ASTRO's English-speaker fans are more attracted to how English lyrics are implemented, even some express interest for ASTRO to release an English song. Also, the disparity between each song's positive and negative comment percentages decreases in number in each new release. It shows the improvement of them in practicing English in their songs as more audiences start to pay attention to it.