Indonesian Twitter Users’ Language Attitude Towards English - Indonesian Code-mixing

Social media users need language as a tool to communicate, both in verbal and written form. Over time, the language used in social media also shows some significant growth. Talking about language in social media, there is an interesting phenomenon recently on social media that can be reached within the internet, namely the code-mixing of Indonesian and English (which in Indonesian are also popularly known as Jaksel language, Keminggris , Kemlondo , Indoglish , Englonesian , or Bahasa Gado-gado ). This study is aimed to find out the attitudes of 500 Indonesian Twitter users towards English-Indonesian code-mixing with a questionnaire. This study uses a descriptive-quantitative approach. For the result of each aspect, it shows a quite positive indication: Cognitive aspect 2.92, Affective aspect 2.75, and Conative 2.79. If the three aspects being summed up, also indicated a quite positive attitude towards the code-mixing with point or score 2.82. The attitudes come from many reasons and factors, such as the need for understandability in communication, language prestige, prepossessing or attracting attention, confidence, and language aptitude.


Introduction
As a result of advances in science, technology, and culture, the use of foreign languages is increasingly entering the daily lives of Indonesian people, especially young people.So many young people nowadays are exposed to new things on the 51 internet already.Technology advances, according to Widawati (2018), also have made the language progress and been promoted very rapidly.Up to 2015, according to Kemkominfo (Indonesia Ministry of Communication and Information Technology), internet users in Indonesia reached 80 million people, an increase of as much as 300% from the last 5 years.Until the following years, internet users in Indonesia had reached 130 to 150 million people.As a matter of fact, 60 million of them had accessed the internet via mobile.Widawati (2018), later said that this is a sign of an extraordinary level of productivity in using language.
Proficiency in English is starting to become an advantage or even a necessity in the modern era.We can see young parents are quickly recognizing this.Over time, the language used in social media also shows some significant growth.Talking about language in general and English on social media, there is an interesting phenomenon recently on social media that can be reached within the internet, which is Jaksel speech style, a linguistic phenomenon where it is namely the codemixing of Indonesian and English (which in Indonesian are also popularly known as Keminggris, Kemlondo, Indoglish, Englonesian, or Bahasa Gado-gado).The use of the language is being discussed, especially on Twitter.Before the Jaksel language variety became a conversation on other social media such as Instagram and Youtube, the Jaksel language variety was already popular on Twitter.This phenomenon came up around September 2018 and it continued spreading quickly on the internet (Rusydah, 2020).Poernamasari (2019) stated the mixing of the two languages is currently considered to be representing the modernization process through social media, and this speech style is commonly used by young people through their daily social activities, either at school, at work environments, when they hang out with friends, or through their social media accounts.
The usage of English mixed with Indonesian or vice versa does not only occur in these platforms and activities, but it happens universally, namely in several other regions in Indonesia (Poernamasari, 2019).Ivan Lanin, one of the writers and language activists on Twitter in an interview with one of the online media, Kompas Cyber Media (2018), states that this phenomenon has been going on for a long time and everywhere, he also stated that this is not a seasonal phenomenon that has just happened currently--or this is the first time it has happened, otherwise its existence will still continue even though jokes or issues, and Jaksel language are no longer widely discussed.
These days, there are many kinds of social networks on the internet.Social networks' users on the internet generally use Twitter as a place to make jokes and to discuss certain social phenomena, one of which is the linguistic phenomenon of speakers of many languages.Prabowo (2018) wrote that recently there has been a lot of bullying on social media for those who mix Indonesian and English.Twitter (or any other social media platforms) is considered as a free and unlimitedaccessed social media, it provides a comfortable space for some people to have free speech, yet also a free media to offend others, even though they have their own regulation, where the violation can be reported and followed-up to certain extents.Poernamasari (2019) furthermore said that mixed language like this might be a symptom that infects many people on Twitter, regardless of the many jokes about a mixture of Indonesian and English which are related to people with better economic conditions or higher education, who have this tendency in everyday conversation.
This kind of phenomenon is inevitably the course of stereotypes.The inherent use of this language seems to give an affirmation to an identity (Poernamasari, 2019).In the end, English is just a language, like race, religion, sexuality, or even haircut style, it is just a human variable.O'Neill and Massini-Cagliari, (2019) stated that variables are quite sensitive to being associated with some type of social meaning or stereotyping, independently of any inherent or intrinsic qualities of the variable.They also stated that such stereotyping can be the cause of linguistic prejudice and even discrimination, which can result in social exclusion and have serious impact for engagement of individuals with the education system and the establishment and thus pose an economic and developmental challenge.
Despite the fuss over the widespread use of mixed language and how people react to these issues and phenomena, Betawi culturalist, Ridwan Saidi, as quoted by Friana (2018) on TirtoID, views the mixed language phenomenon as a creative behaviour.According to him, this bilingual pattern in language is not destructive.On the other hand, Prabowo (2018) on TirtoID quoted a clinical psychologist, Kasandra Putranto, who stated that if the Jaksel language variety is used as a method of learning English, it should not be viewed as something negative, according to him, every era has its own style of language.Prabowo (2018) on TirtoID also quoted what Ikwan Setiawan stated, a lecturer of the Faculty of Cultural Sciences, Universitas Negeri Jember (UNEJ), that the phenomenon of young people using mixed languages is normal behaviour and academically is not a problem.
It becomes very interesting to see when the function of language as it can unify nations and other positive values, it also begins to change in a negative direction.Language is then used as a tool that can corner, offend, and oppress other people through interactions in person or through social media on the internet.Stereotyping a language and any beliefs towards a language is a discourse in attitude towards a language and it then becomes a sub-topic in sociolinguistics.For example, judging a language as ugly, beautiful, or harsh and the speakers of the language as rude, rich, sophisticated, or educated (Schu ppert et al., 2015).According to Dragojevic (2017), every person surely has a different opinion about certain different things, including opinions about a language.He stated that language attitudes are appraising reactions or responses to different language varieties.They can be the opinion, ideas, and prejudice that the speakers would have with respect to a language.They reflect two sequential cognitive processes: social categorization and stereotyping.
The writer becomes excited to explore attitudes, as the writer himself has seen so many biases and cases towards multilingualism, especially on the internet.In this case, the writer also has been using Twitter for a decade and seeing so many linguistic phenomena where some people would be so angry and bitter by only seeing other people's writings on social media that are written in certain varieties or other foreign languages.But some others are very supportive.So that it can refer to negative or positive attitude, integrative or instrumental attitude, and how language attitude can be measured with some methods.
Recent studies in language attitude, like Rossi and Saneleuterio (2016), found that language attitudes play an important role in bilingualism and second language learning.Among the many ways, we often hear foreign languages being used and the language mixing becomes inevitable.As code-switching can be felt as a problem or as a result of being open-minded in language learning.The paper analyses the responses of a questionnaire by a sample of 53 bilingual subjects.Most of them are English and Spanish speakers, and it is relevant when it comes to the correlation between the number of languages a person knows and the language prejudice of someone.When it comes to code-switching and code-mixing, Abdurahman (2020) found that the variety of English and Indonesian which is grammatically independent, i.e., alternation, is perceived as more socially attractive, while the standardized English is perceived as more intelligent.The conclusion from his study is that the ideological and social factors may affect people's perception and overall attitude towards the use of English in Indonesia and the mix or switch of these both languages.Also, Abdurahman, Gandana, and Novianti (2018) explored Indonesian university students and faculty members' attitudes towards the use of English in both face-to-face and virtual contexts.The results show that there were mixed attitudes towards English among the respondents.This study suggests that while virtual domains can provide a space for learning and practicing English, a beneficial utilization of the language ultimately depends on how English language learning is planned and designed.
Last, research by Naveed (2014), he used a questionnaire to collect the data from 200 students in four Pakistani colleges and universities to find out the students' attitudes towards both educators and students' code-switching, as well as reasons for performing and not performing to code-switching in EFL contexts.
The results indicated positive attitudes of students towards using the target language by themselves and their educators.Meanwhile, using Urdu was perceived beneficial for expressing ideas, explaining new vocabulary words, and optimizing the learners' chance to improve their English proficiency.
From the studies above, there are similarities and differences to the current research, namely that all of them focus on investigating and exploring people's attitudes towards language and language variations.Several of them are investigating the attitudes towards phenomena in multilingualism, one of which is code-mixing and code-switching.However, those studies are different from the present study in terms of the domain of the community that participated.Exploring language attitudes, also means that it is possible to be able to explore the reasons why someone chooses a language (code) or even variations of languages, including code-mix and code-switching.Otherwise, this study will find out more about attitudes towards code-mixing--which is English-Indonesian code-mixing--in the Twitter domain.The participants are Twitter users across Indonesia.Another difference is the data collecting method that will be used.
The writer finds this research interesting and useful to explore where the writer himself has not seen many similar studies.The writer sees not so much research on language attitudes on social media in Indonesia.The attitude of social media users in Indonesia towards the use of English mixed into Indonesian is barely unknown.Therefore, sociolinguistic research on language attitudes towards the use of English should be explored.This research will focus on studying the language attitudes of Indonesian Twitter users towards English-Indonesian codemixing.The writer chooses this title to find out about language attitudes on the internet, specifically on Twitter.Based on the background above, this study is undertaken to answer the following questions: 1) What are Indonesian Twitter users' language attitudes towards English-Indonesian code-mixing?

Method
In this study, the writer used a descriptive-quantitative approach.The population of this study were mainly Indonesian Twitter users.There are so many Twitter users nowadays.Prihadi (2015, cited in Nugraheni, 2017) emphasized that in 2015, the number of Twitter users in Indonesia is estimated to have reached 50 million.Since there are limitations in time, energy, and funds, the writer deliberately limited the research subjects and used samples instead.For the sample, this study used non-probability sampling.The writer collected 500 respondents with self-selection sampling technique (or also known as voluntary sampling technique).The participants will be the ones who perform English-Indonesian code-mixing.
This study used a questionnaire to obtain the data about the attitudes of Twitter users, where the items were structured in the form of open-ended and closed-ended questions.The questionnaire was created using Google Forms.It was then distributed and broadcasted on Twitter.The items in the questionnaire in this study used and adapted questionnaire items from Al-Qaysi (2016) who had conducted research about attitude towards code-switching previously, but in this study, the term "code-switching" is changed to "code-mixing" to adapt to the topic of this research.Since Indonesian is the national language, the questionnaire was translated into Indonesian to avoid misunderstanding and to assist the participants in selecting the proper options and context.
The questionnaire survey includes 18 items that are categorized into three sections.The first section includes 2 items representing the respondents' demographic data: gender and age.The second section includes 1 question to represent the respondents' usage of code-mixing in daily life and social media interaction.The last section consists of 14 items of statement investigating the respondents' attitude towards the code-mixing.
For the data analysis technique in this study, Likert scale is used related to the respondents' language attitudes.Every answer is given a score with a range from one to four, with a range from 'strongly disagree', 'disagree', 'agree', and 'strongly agree'.The highest score is 4 (four), while the lowest score is 1 (one).The highest score indicates a high positive attitude, then the lowest score indicates the other way around.The responses were analyzed in Microsoft Excel.The writer calculated the average of the answers, calculating the percentage, and classifying the dominant percentage.Conclusions were drawn by calculating the average points obtained from the respondents.To calculate the percentage of the answers, the writer used this pattern:  = (  ℎ   100)

Results
As for the characteristics of the respondents, they will be described by gender, age, and usage frequency or how they perform the code-mixing in daily communication, especially in social media (Twitter).

Gender of the Respondents
From the chart 1, we can see and conclude that women are the most participated in this research with a total of 331 respondents or around 66.2%.Then followed by 119 male respondents (23.9%) and undecided with 50 respondents (10%).

Age of the Respondents
From this Chart 2, we can see that the dominant respondents come from the range of age 17 to 25 with a total of 321 respondents (64.2%).Then followed by 26-35 with 112 respondents (22.4%).By this, we can conclude that the respondents from late teenagers to young adults are the ones that participate the most in this research.

Usage Frequency of English-Indonesian Code-mixing
From Chart 3, we can see that the respondents tend to perform the English-Indonesian code-mixing in this research.There are 287 respondents (57.4%) who choose "Sometimes" and "Always" with 120 respondents (24%).We can conclude that the respondents that participate in this research, have high usage of the English-Indonesian code-mixing.) respondents that chose Agree and 169 (33.8%) respondents chose Strongly Agree for this first item.Here, we can conclude that the respondents (78%) agree that English-Indonesian code-mixing can help to enhance their communication skills.Next, there are 185 respondents (37.0%) on Agree and 227 (45.4%) on Strongly Agree for item 2. Mainly, we can conclude that the respondents (81.9%) agree on the code-mixing to develop their language skills.
For item 3, showed that there are 157 respondents (31.4%) on Strongly Disagree, 167 respondents (33.4%) on Disagree, 132 (26.4%) respondents on Agree, and 44 respondents (8.8%) on Strongly Agree.The distribution of the response is quite fair.But we can still see that most of the respondents (64.8%) do not agree that English-Indonesian code-mixing can show them off as a welleducated person, while the rest believe that the code-mixing shows them that they are.
On Item 4, asked about how the code-mixing is still related to their perceived language prestige where they think that the code-mixing could make them feel more prestigious.This item revealed that there were 220 respondents (44.0%) on Strongly Disagree and 185 respondents (37.0%) on Disagree.The distribution of the responses is quite extremely polarized on negative answers.We can see that the respondents (81.0%) do not agree that the code-mixing can make them look prestigious.
Item 5, revealed that 40.0% of them stand on the Agree side and 26.0% on the Strongly Agree.We can conclude that 66.0% of them agree that the code-mixing can make them understand better.Then, item 6, revealed that 37.0% of them agree with it and 47.6% of them strongly agree.Here, we can conclude that 84.6% of them agree that the code-mixing can help to convey new words easily.For item 7, revealed that 30.8% of the respondents agree and 50.6% strongly agree with it.This can mean that most of them (81.4%)agree that the code-mixing can help them to practice the second language.
Item 8, asked about how the code-mixing can help the respondents learn new words from friends or teachers (school or university).This item indicated that 39.4% of the respondents agree and 43.8% of them strongly agree with the notion.This means that most of them (83.2%)agree that English-Indonesian code-mixing can help them to learn or memorize new words.
Last, item 9, showed that 25.2% of the respondents agree and 65.0% strongly agree with the statement.From this, we can conclude that 90.2% of the respondents agree that the usage of English can help them express ideas that they cannot do in Indonesian.Item 10, has 34.8% of the respondents who agree and 32.2% who strongly agree with the statement.We can conclude that 67.0% of the respondents agree that the code-mixing might attract their attention.

Affective Component
Item 11, has 34.0% of the respondents who agree and 21.8% who strongly agree with the statement.On the other hand, there is an almost equal number of responses to this statement, with 30.4% of respondents disagree and 13.8% of respondents strongly disagree.Mainly, we can conclude that 55.8% of the respondents are the dominant one for this statement where they mostly agree.On item 12, 29.0% of the respondents agree on this statement and 42.4% strongly agree with the statement.It can be concluded that the respondents (71.4%) tend to agree that they do mix Indonesian with English due to the lack of Indonesian equivalents.
Item 13, has 32.2% of the respondents who choose to agree with and 29.4% of them chose to strongly agree with the statement.To sum up, the respondents (61.6%) consider using English-Indonesian code-mixing if there are complex words in their native language.
Last, item 14, there are 54.2% of respondents who admit that they do, with the following split: 32.6% of them agree and 21.6% chose to strongly agree.Here, it can be concluded that only about half of the respondents admitted that they mix English and Indonesian in discussing lessons, lectures, and exams.

Average Result
For the whole result of score, it shows that the respondents (as Twitter users) indicate they tend to agree and have positive attitude towards the English-Indonesian code-mixing with the details are as follows: Strongly Agree (4) has 2365 count score (33.8%) from all 14 items of the questionnaire and followed by Agree (3) with 2285 count score (32.6%), while Disagree (2) only has a count of 1410 score (20.1%), the Strongly Disagree (1) has count 940 score (13.4%).
For the result of each aspect, each aspect shows a quite positive indication (cognitive aspect 2.92, affective aspect 2.75, and conative 2.79).For the result of three aspects being summed up, the attitude has a mean score 2.82.

Discussion
This research sought to find out Indonesian Twitter users' language attitude towards English-Indonesian code-mixing.From the age of the respondents, we can see that the respondents from late teenagers to young adults are the ones that participate the most.The young generations are digital natives, as they grew up surrounded by digital technologies, they are native speakers of the digital language of computers, video games, and also the Internet (Prensky, as cited in Jongbloed-Faber et al., 2016).We Are Social (wearesocial.com, in Ahmad and Nurhidaya, 2020) found that there was an increase in the use of social media compared to 2018 and its use was dominated by young people in Indonesia's generation Y (millennials) and Z, namely between 18-34 years old.
The overall mean indicated that the users are positive towards the codemixing, yet if we break down more on the specific statements, there are few statements that are not agreed by the respondents in this finding.There are also items that get an almost balanced response.
We can start discussing the findings that are related to communication.They showed how English-Indonesian code-mixing might help the respondents to communicate better, either about improving communication skills, letting them understand more easily, conveying new words regarding current hot issues, the lack of equivalent words in source language to express ideas more precisely or due to the lack of equivalent words in native language or mother tongue.
These findings of this research have several items that related to communication performance with a high percentage of positive values.Thus, this research supports various studies conducted in the sociolinguistics domain, especially language alternation.Mujiono et al. (2017), conducting research about how code-mixing can be a communication strategy performed by Outbound Call (OBC) Center Agents.In their research, they found out that English-Indonesian code-mixing happened in different situations in order to appreciate the customer, to attract customer's attention or to persuade the customer, and to explain about the weakness of the products.This is in-line with what Kim (2006) described about the reasons or the social factors that influence the usage of code-mixing, such as participant roles and the relationship, situational factors, message-intrinsic factors, and as well as attitudes, dominance, and security.This is also in line with Bhatia and Ritchie (2008), who described backgrounds and relationships of participants, topic or content, and when and where a speech act occurs can be the triggers for people to mix their languages.
Furthermore, nowadays we know we have sophisticated mobile phones with decent and accessible internet connection for all.The internet makes it easy for anybody to access English-language based content, across countries, which means it also opens opportunities for people to be exposed to English.We also know that English has become the language of instruction in social media.Then, this can be the cause of why English and the mixing is inevitably used in various social media, especially Twitter.
Twitter only allows the users to write status (called as tweet) with limited characters.This way makes Indonesians consider switching some words in Indonesian language with words in English language or vice-versa.This is in-line with Abdurahman et al. (2018), where the study found out that the code-mixing enabled participants to write a shorter status on Twitter.Alternating with English can also be the reason why this language is chosen when it comes to avoiding complexity, especially conveying new words they have just recognized, that are related to technology, politics, or medical, such as gadget, social justice warrior (SJW), and swab test.
We can also begin to talk about the items in cognitive components that talked about language prestige.They somehow get polarised in a negative direction.The users largely do not consider the English language as EFL in Indonesia as an education benchmark.The majority think that knowing English and mixing English words is not a guarantee to be perceived as a well-educated person.This is in-line with a study by Abdurahman (2020), where only standardised English is found to be perceived as intelligent, not the mixed variety.On the other hand, the codemixing English-Indonesian is found to be perceived as socially attractive instead.
The users also do not consider that English-Indonesian code-mixing can show them as prestigious person.This result is similar to Al-Qaysi (2016) and Wirojwaranurak (2017) where the participants, fairly half or less, disagree with English and the mixing variety on certain prestige or pride.At the same time, this finding does not support several studies about Jaksel, Keminggris, or Indoglish variety as a form of English-Indonesian code-mixing, that has proven the existence of language prestige towards English, especially in urban areas or cities will increase the tendencies to code-mix both languages (Rakhmawati et al., 2016;Saddhono and Sulaksono, 2018;Oktavia, 2019;Rachman, 2019).In this study, the writer assumes that this can happen since Twitter is seen as a very informal social media.Also, the English language has been considered as a common language on the internet.There is likely for the mixed variety to be perceived so commonly, especially on social media since code-mixing of Indonesian and any local languages such as Javanese, Bataknese, Sundanese, or Bugis also happened on Twitter.Everyone also can write as they please.People can also write as freely as possible, with their own style of writing and expression, where they feel like no need to worry too much about the language choice and how other users perceive their identity on the internet especially since we can make a profile on Twitter where we do not need to upload our identity of who we are.Twitter then remains as a personal blog.
For the discussion of the items in the affective domain, we can conclude that they have a fairly high value where the respondents mostly agree that English-Indonesian code-mixing can attract their attention and also, they feel confident towards it.Code-switching in a study by Mukti and Muljani (2016) shows that students are more interested and pay attention to teachers when explaining lessons in class when the teachers perform the code-mixing.Then, code-mixing in a study by Leung (2010) about code-mixing in advertisements found that codemixing alone can attract the attention of readers and listeners, let alone the audience who has a higher educational background.Regarding the educational background, this is also in line with research by Itmeizeh and Badah (2021) where they found people with higher educational backgrounds who feel confident with code-mixing, which is also influenced by one's language competence or acquisition.
Last, the statements that have the nexus of language attitudes to language learning.This topic becomes a cross-over with language aptitude.According to Stansfield (1989), language aptitude can refer to prediction of how well a person's ability to learn a foreign language in a given period of time and under given conditions.These items show very high indication that the respondents--when it comes to English--consider the code-mix as a process of learning, improving language skills, and practicing the second language and English language at once.Having positive attitudes towards a language, often leads to a significant increase in language skills.Gardner (2014) stated there are two types of motivations related to language attitude, namely integrative and instrumental motivation.The statements mentioned about learning language are in parts of integrative and instrumental motivation.Integrative orientation deals with the reason suggesting that an individual learned a particular language in order to learn about or interact with a community or other people or users on the internet who happened to use the same variety.Surely this will help the process of language acquisition.Jismulatif (2018) stated that successful language learning, especially English, is heavily influenced by the positive attitude towards the language itself, besides that motivation also has a very important role in realising success in learning a language.He also stated that the attitudes towards language learning itself contributes to the success rate.This is in line with study by Oroujlou and Vahedi (2011), where they also stated that in learning a language, it often helps to have a positive attitude towards that language.Riyanto et al. (2015) also found that one of the factors that influence students' vocabulary understanding is the attitude towards language learning.research also seems to support Henni (2017) where in her study, she found out that positive views towards code-switching seems to have a fruitful role in the learning process.Last, this supports the study by Omar and Ilyas (2018).They found out that attitude towards code-switching affected learners' academic performance since the learners' attitude towards Arabic and English language contributed to their learning and knowledge acquisition.

Conclusion
Based on the research finding, the Indonesian Twitter users indicate that they have quite a positive attitude towards the code-mixing of English-Indonesian.Each aspect of attitude also shows quite a positive attitude towards the code-mixing of English-Indonesian.The users mainly agree that the English-Indonesia codemixing can enhance the understandability in communication.Regarding language prestige, the users mainly disagree that the English-Indonesia code-mixing makes them feel more prestigious.The users agree that the English-Indonesian codemixing can draw their attention.They also tend to agree that they feel confident and comfortable towards code-mixing.The users when it comes to English mainly consider the English-Indonesian code-mixing as a process of learning, improving language skills, and practicing the second language and English language (as EFL in Indonesia) at once.This research focuses on the attitudes of Indonesia Twitter users towards English-Indonesian code-mixing.This study does not really dig further into the respondents' language proficiency.Regarding language proficiency, this study does not aim to find how language proficiency might influence the attitude or vice-versa.This study also does not find the language choice (reasons or factors) and the language attitude towards each language or variety to be compared, whether it is in daily interaction or in Twitter per se.So, for future studies, the writer hopes for the other researchers to conduct the same topics with more participants in the different domains or specific groups of population (i.e the people who do not perform English-Indonesian code-mixing) to gain different findings from different perspectives and comparing between Indonesian, English, and English-Indonesian code-mixing.Besides that, it is also better if the next researchers can link the study to the latest sociolinguistic theories to ensure the research's recency.Furthermore, future researchers can use more various data collecting techniques or methods and instruments, such as interviews and focus group discussion.Last, the writer also really hopes for the next researchers to be able to construct more precise items with larger concepts and compare the attitudes of groups belonging to different generations, gender, educational background, and occupations to make the data richer.

57
Chart 3 Usage Frequency pf the Respondents' Code-mixing

Table 1 Cognitive
Aspect of Respondents