*Corresponding Author P-ISSN: 1978-8118 E-ISSN: 2460-710X 207 Lingua Cultura, 15(2), December 2021, 207-214 DOI: 10.21512/lc.v15i2.7335 ON GOOGLE TRANSLATE: STUDENTS’ AND LECTURERS’ PERCEPTION OF THE ENGLISH TRANSLATION OF INDONESIAN SCHOLARLY ARTICLES Menik Winiharti1*; Syihabuddin2; Dadang Sudana3 1,2,3Department of Linguistics, School of Post Graduates, Indonesia University of Education Jl. Dr. Setiabudi 229, Bandung 40154, Indonesia 1English Department, Faculty of Humanities, Bina Nusantara University Jl. Kemanggisan Illir III No. 45, Palmerah, Jakarta 11480, Indonesia 1menikw@upi.edu; 2syihabuddin@upi.edu; 3dsudana@upi.edu Received: 11th May 2021/Revised: 13th October 2021/Accepted: 13th October 2021 How to Cite: Winiharti, M., Syihabuddin, & Sudana, D. (2021). On google translate: Students’ and lecturers’ perception of the English translation of Indonesian scholarly articles. Lingua Cultura, 15(2), 207-214. https://doi.org/10.21512/lc.v15i2.7335 ABSTRACT The research investigated the translation results of Google Translate based on the users’ perceptions. It was aimed at describing the users’ frequency in using Google Translate web, finding the users’ perceptions on the acceptability and the readability of Google Translation in translating scholarly articles from Indonesian into English, as well as finding whether students and lecturers have the same perceptions regarding these two criteria. The data were collected from scholarly articles written in Indonesian then translated into English using Google Translate web. Then a survey regarding this translation was distributed to users; they were students and lecturers of Computer Science/Information Technology/Information System. The analysis was conducted with regard to the acceptability and the readability of the translations using a rating scale of translation assessment. The findings suggest that more than half of the participants often use Google Translate Web, which means that the academics are part of the users of Google Translate. However, students and lecturers have a rather different perception of the results of Google Translate. Students consider Google Translate quite acceptable and readable, while lecturers view Google Translate as rather acceptable and moderately readable. In addition, the findings indicate that, to some extent, Google Translate still translates the Indonesian text into English literally. Keywords: Google Translate, translation assessment, users’ perception, translation acceptability, translation readability INTRODUCTION Google Translate is one web-based machine translation that has been developing quite rapidly. It says that this machine has been improving quite significantly. Several pieces of research (Brahmana, Sofyan, & Putri, 2020; Habeeb, 2019; Khosravizadeh & Pashmforoosh, 2011) have shown the pros and cons around the quality of Google Translate. From one viewpoint, Google Translate is considered an essential tool for individuals to connect with others of different languages, yet it does not imply that the machine is consistently impeccable in rendering all languages. Khosravizadeh and Pashmforoosh (2011) have pointed out that Google Translate has some limitations despite the fact that it is beneficial for helping language users around the globe. In this case, according to them, the involvement of humans in the editing process is yet required for a better version of translation (Khosravizadeh & Pashmforoosh, 2011). Habeeb (2019) has also stated that even though Google Translate has several advantages, it has drawbacks such as inaccurate output and no proofread tools. A similar result is shown by Brahmana, Sofyan, and Putri (2020), who have found that the major problem of Google Translate is its inaccuracy and inappropriateness of meaning and its inaccuracy in the structure. Furthermore, regarding language learning, 208 LINGUA CULTURA, Vol. 15 No. 2, December 2021, 207-214 Hoi (2020) has suggested that even though machine translation has been evolving and able to increase humans’ efficiency, much dependence on this machine should be avoided. The impact toward the users relies on themselves. Thus, it is suggested that users still need human translators because if they only depend on the machine, they would only learn simple things and cannot develop their language ability, which would influence the way they communicate (Hoi, 2020). Furthermore, Alsalem (2019) has conveyed a stronger statement that humans do not need to over-rely on Google Translate, especially in the editing process. For students to develop their translation skills, they need to be guided and experience the learning in the editing stage; thus, they are not suggested too much depending on Google Translate. Even in the course of translations for undergraduates, Siregar et al. (2020) have suggested that students have to be aware of the importance of following the procedures, the cultures- different contained in languages, the use of translation tools in finding the best equivalence, preparing the best version by act as self-editor, and benefit of this activity in enhancing the students’ ability in English. This implies that using a translation tool or machine translation alone is not suggested to provide a good translation result. There are other procedures incorporating the translation tool. For this case, even though Brahmana, Sofyan, and Putri (2020) have stated that Google Translate can be used as a learning media in translation, they have also suggested that students make self-corrections and consult dictionaries to confirm the correct meaning of doubted words then choose the best option by considering the context. On the other hand, Valijärvi and Tarsoly (2019) have viewed Google Translate more positively, especially with regard to foreign-language learning. By incorporating Google Translate as a tool in the learning process, students who learn foreign languages (in their research, it is Finnish and Hungarian) become more analytical and proficient, which has encouraged them to learn the languages autonomously, which in turn would increase their confidence. Correspondingly, Karjo and Metta (2019) have observed that undergraduate students are quite ‘prolific’ in using Google Translate, especially finding the equivalence of difficult words to complete their essay writing. Even though they state that students may use Google Translate as a translation tool, students should not be too dependent on the machine. Thus, the lecturers should provide proper guidance in order to provide a better result in translation. These researches (Alsalem, 2019; Brahmana, Sofyan, & Putri, 2020; Habeeb, 2019; Hoi, 2020; Karjo & Metta, 2019; Khosravizadeh & Pashmforoosh, 2011; Siregar et al., 2020; Valijärvi & Tarsoly, 2019) imply that Google Translate has been widely used by people around the world, including those who are involving in the academic fields. Therefore, the needs to evaluate the result of Google Translate within the academic environment are important. Regarding such evaluation, there is a parameter of translation result which involves accuracy, naturalness, and clarity (Larson, 1998). Assessing the translation accuracy generated by Google Translate, Aiken (2019) has reevaluated his previous findings that suggested translations between European languages are usually good, while those involving Asian languages are often relatively poor. He compares 51 languages, with Indonesian being included in the list of comparisons. His recent findings have suggested that based on the measurement using BiLingual Evaluation Understudy (BLEU) scores, Indonesian has decreased 14,6%, but there is a 34% improvement of the overall 51 languages. While assessing the translation quality generally focuses on accuracy, in fact, it should involve other criteria that concern the perception of the audience who become the end-users of the translations. These other criteria are acceptability and readability, as proposed by Nababan, Nuraeni, and Sumardiono (2012). Similarly, Larson (1998) has also suggested clarity and naturalness, in addition to accuracy for translation evaluation. With regard to these criteria of translation assessment, research conducted by Kartika and Priyatmojo (2018) shows that Google Translate is less accurate, less acceptable, and less readable in translating newspaper texts from English into Indonesian. A somewhat similar result is shown by Setiyadi et al. (2020), who has found that Google Translate is less clear and quite unnatural in translating cultural-specific concept from English into Indonesian. In addition to these backgrounds, students and lecturers in Indonesia are required to publish academic articles in English to increase the university’s rank on the international scale to obtain more international recognition. Unfortunately, their writing can be hindered because they lack English writing skills (Arsyad et al., 2019; Hartono, Arjanggi, & Praptawati, 2019). This is where the translation plays an important role, and in this context, Google Translate is quite desirable to use because it is easy to access, and the results occur instantly. The present research is conducted because studies on the translation from Indonesian into English are not likely much conducted, specifically regarding the users’ point of view on Google Translate results. For this case, college students’ and lecturers’ perceptions of Google Translate are crucial, assuming that to some extent, they use Google Translate as a tool to translate some texts within their tasks. Therefore, the present research has three goals to achieve. The first is to describe the users’ (students and lecturers of Computer Science/Information Technology/Information System) frequency in using Google Translate Web. The second is to find their perception of the acceptability and the readability of Google Translate in translating scholarly articles from Indonesian into English. The third is to find whether students and lecturers have the same perception regarding the two criteria of translation assessment. When these goals are achieved, the findings may fill in the gap in the area of studies, in this case, is Google Translate quality perceived by the 209On Google Translate:.... (Menik Winiharti, et al.) academics in Indonesia. It may also enrich the previous ones and confirm those with related topics. Nababan, Nuraeni, and Sumardiono (2012) have suggested that a good quality of a translation must meet three aspects; aspect of accuracy, acceptability, and readability. The concept of accuracy refers to whether the target language (TL) and source language (SL) are equal. Being equal means that the message conveyed in the TL is faithful to that in the SL. The aspect of acceptability deals with the norms and rules in the TL; whether the translation has applied the rules and norms in the TL. This includes linguistic rules and socio-cultural rules. For the third aspect, readability, Nababan, Nuraeni, and Sumardiono (2012) do not define this aspect; instead, they describe that determining the level of readability is not easy as it is determined by the readers’ knowledge. Longman dictionary defines readability as being ‘interesting and enjoyable to read, and easy to understand’ or ‘writing or print that is readable is clear and easy to read’. Based on this definition, the aspect of readability can be referred to whether the target audience can easily read the translation and that the translation is precise. Meanwhile, Larson (1998) has proposed that accuracy, clarity, and naturalness should become the features to be considered when a translation is evaluated. According to her, a translation is said to be accurate when the message of the SL is equally conveyed in the TL. There is no addition or deletion of information. Then a translation is said to be clear when the message is easily understood by the TL readers. In other words, the translation communicates well to the audience who uses it. Finally, a translation is natural when using forms that are familiar in the TL. The forms here must be in accordance with the TL grammatical rules; thus, the style is idiomatic, and the translation does not sound strange (Larson, 1998). It can be said that the criteria of translation evaluation or assessment proposed by Nababan, Nuraeni, and Sumardiono (2012) and that by Larson (1998) are basically similar. Table 1 describes the parameters of a good translation proposed by these scholars. Table 1 Parameters of a Good Translation by Nababan, Nuraeni, and Sumardiono (2012) and Larson (1998) Nababan, Nuraeni, and Sumardiono (2012) Larson (1998) Description Accuracy Accuracy The message/meaning of a translation is equal/ faithful; no distortion in meaning Acceptability Naturalness The forms use the TL’s rules and norms; socio-culturally and linguistically Nababan, Nuraeni, and Sumardiono (2012) Larson (1998) Description Readability Clarity The translation is clear and easily read/ understood Furthermore, Nababan, Nuraeni, and Sumardiono (2012) have proposed the rating scales of their three parameters of a good translation, each of which is graded from 1 to 3, with 1 being the least and 3 being the best. However, Larson (1998) does not propose the rating scales to evaluate the translation. She only discusses how to rate the translation qualitatively. METHODS The research applies a qualitative descriptive approach as it tries to describe the users’ perception of the quality of Google Translate regarding the two aspects of evaluation: acceptability and readability. The data are of Indonesian scholarly articles. They are written in Indonesian, by Indonesian scholars, and published in Indonesian journals, at least nationally accredited Sinta 3. They are collected and downloaded from national journals, which can be accessed through a web page: https://www.neliti.com/id/journals. The area of the articles is sorted to those of Computer Sciences, Information System, and Information Technology. There are five articles downloaded. They are selected randomly but sorted into three criteria: published after 2015, written in the Indonesian language, and written by Indonesian scholars. Since twenty sentences are used as the Indonesian sources, four sentences from each article are collected. These sentences are limited to those written in the conclusion section, assuming that this part is generally the authors’ own language, which summarizes the whole result and discussion. The twenty Indonesian sentences are then inputted into Google Translate web https://translate. google.com/ to be translated into English. Afterward, a survey containing this translation is created using Google Forms. Thus there are twenty English- translated sentences in the survey, examining the acceptability and readability of the translations. The present research would like to apply the parameters proposed by Nababan, Nuraeni, and Sumardiono (2012) to assess these aspects. Because the research emphasizes the users’ perception toward the result of Google Translate, it chooses to use two criteria as they have proposed: acceptability (Table 2) and readability (Table 3). Furthermore, as they have only three rating scales for assessing a translation, the research needs to determine a broader range of rating scales. Therefore, Table 1 Parameters of a Good Translation by Nababan, Nuraeni, and Sumardiono (2012) and Larson (1998) (Continued) 210 LINGUA CULTURA, Vol. 15 No. 2, December 2021, 207-214 the research modifies theirs so that the scales become 5, with 5 being the highest scale (the best) and 1 being the lowest (the worst). For the criteria of acceptability, it is adopted from Nababan, Nuraeni, and Sumardiono (2012) with the addition of two scales. Thus, there are five grades as the rating scales, with 5 being the most acceptable and 1 being the least acceptable translation. Table 2 The Present Research’s Acceptability Scales (Modified based on Nababan, Nuraeni, and Sumardiono, 2012) Grades Description 5 Translation is natural; the technical terms used are common and are familiar to the readers; phrases, clauses and sentences used are in accordance with Indonesian rules. 4 Generally, the translations are natural; however, there is a slight problem on the use of technical terms, or a slight grammatical error occurs. 3 Translation is quite natural, but there are many unusual uses of technical terms or there are many grammatical errors. 2 Translation is not natural, the translation sounds like a translated work. There are quite a lot of technical terms that are not common; the use of phrases, clauses and sentences are not in accordance with the rules of the target language. 1 Translation is very unnatural, or the translation sounds very much like a translated work; the technical terms used are not common and are not familiar to the readers; phrases, clauses and sentences used are not in accordance with the rules of the target language. The second criteria, readability, is basically modified from Nababan, Nuraeni, and Sumardiono’s readability (2012). Originally, it contains three scales. The research develops the scales into five, with 1 being the least readable and 5 being the most readable. The aspect of readability deals with the ease of the readers in understanding the translation. It involves the use of words, phrases, clauses, and sentences that are easily and clearly understood by the TL’s readers. Furthermore, the survey is then distributed to users. They are twenty-five undergraduate students and three lecturers of Computer Science/ Information Technology/Information System. Students’ and lecturers’ perceptions are separately discussed to find if their perception toward Google Translation result is the same. These participants are selected from those having TOEFL scores of more than 500 and good mastery of English. Out of twenty-five students, seventeen have a score above 550, while eight of them have a score of 501-550. Of three lecturers, only one has a score above 550, while two of them have a score of 501-550. The users are also asked about their frequency in using Google Translate, whether they never, sometimes, or often use the machine. This is to confirm that they have ever used Google Translate. Table 3 The Present Research’s Readability Scales (Modified based on Nababan, Nuraeni, and Sumardiono, 2012) Grades Description 5 Clear and intelligible; has no or very few non-standard words, expressions or grammar. Readers can easily understand the words, technical terms, phrases, clauses, sentences or translated texts. 4 Mostly clear and intelligible; contains some non-standard words, expressions or grammar. 3 Generally intelligible; contains many non- standard words, expressions or grammar. In general, readers can easily understand the translation; but certain parts must be read more than once to understand it. 2 Generally unintelligible; contains many non- standard words, expressions or grammar. 1 Unintelligible; dominated by non-standard words, expressions or grammar. Readers are difficult to understand the translation. After all the data are collected (translated texts and users’ perceptions), the analysis is done in two stages. First, it deals with the result of users’ perception regarding the acceptability and readability of Google Translate. This would mainly use Nababan, Nuraeni, and Sumardiono’s (2012) concepts of acceptability and readability of translations modified into the rating scales listed in Table 2 and 3, respectively, with 5 being the highest and 1 being the lowest score. The results of the acceptability and readability of the translations are calculated using simple calculations. Second, the analysis deals with linguistic aspects as an example of discussions when the translated sentences acquire low or high scores in terms of their acceptability and readability. RESULTS AND DISCUSSIONS The first result deals with the frequency of the users – undergraduate students and lecturers – in using Google Translate web. Figure 1 shows the students’ frequency, while Figure 2 displays the lecturers’ frequency using Google Translate web. Figure 1 shows that out of 25 students, 14 students (56%) use Google Translate web often, while 11 students (44%) use the machine sometimes. It can be said that more than half of them use Google Translate web often. It seems that the frequency of the lecturers in using Google Translate web is not entirely different from their students. Out of three lecturers, two (67%) use Google Translate web often, while only one (33%) uses the web sometimes. Nevertheless, the results are shown in Figures 1 and 2 inform that Google Translate indeed has certain users; even undergraduates and 211On Google Translate:.... (Menik Winiharti, et al.) lecturers often use it. Figure 1 Students’ Frequency in Using Google Translate Web Figure 2 Lecturers’ Frequency in Using Google Translate Web The next results deal with those regarding the criteria of acceptability and readability from both kinds of users, undergraduate students and lecturers. First, Table 4 displays the students’ point of view with regard to both assessment criteria on Google Translate, acceptability and readability. Table 4 Average Scores of Students’ Point of View on Acceptability and Readability of Google Translate Sentence Number Average of Acceptability Score Average of Readability Score 1 3,76 4,00 2 3,96 4,16 3 3,44 3,80 4 3,36 3,36 5 3,48 3,52 6 4,16 4,20 7 3,28 3,36 8 3,72 3,76 9 3,60 3,68 10 3,56 3,64 11 4,08 4,20 12 4,12 4,24 13 3,84 4,12 Sentence Number Average of Acceptability Score Average of Readability Score 14 4,12 4,12 15 3,80 4,00 16 3,72 3,72 17 3,72 3,56 18 4,16 4,20 19 3,88 4,04 20 4,04 3,88 Average 3,79 3,88 Max 4,16 4,24 Min 3,28 3,36 Table 4 indicates that the acceptability score of the translations resulting from Google Translate is moderately good or quite acceptable, with an average score of 3,79 (out of 5). While the readability score is a little bit higher, reaching 3,88, though it is still below 4 (out of 5). In other words, Google Translate translations are considered quite readable. Overall, the undergraduate students consider that the English translations of Google Translate are moderately acceptable and quite comprehensible. The following are examples of sentences that need to be discussed as they receive the lowest or the highest scores. Example 1. The sentence refers to sentence 7 in Table 4, which receives the lowest score for both acceptability and readability. Indonesian : Aplikasi sistem pakar “Awas Meningitis!” dapat melakukan proses diagnosis penyakit sesuai data rekomendasi yang didapatkan dari pakar. English : Application of the expert system “Beware Meningitis!” can make the process of disease diagnosis according to the recommendation data obtained from experts. This sentence is considered the least acceptable and readable even though the scores reach 3,28 and 3,36 respectively; thus, it means that students think this is moderately good with regard to the criteria of acceptability and readability. However, Google Translate seems to translate the sentence quite literally. In the English translation given, the noun phrase “proses diagnosis penyakit sesuai data rekomendasi yang didapatkan dari pakar” in fact does not sound quite natural in English. It is supposed to be “the process of diagnosing diseases based on the data recommended by the experts.” Thus, from the researchers’ point of view, it is scored 2 for acceptability and 3 for readability. It is readable to some extent but not natural in English. Table 4 Average Scores of Students’ Point of View on Acceptability and Readability of Google Translate (Continued) 212 LINGUA CULTURA, Vol. 15 No. 2, December 2021, 207-214 Example 2. This example is of sentence 12 in Table 4, which receives the highest score regarding its readability. The score of its acceptability is quite high, second highest in fact, even though it does not reach the highest one. Indonesian : Data yang digunakan dalam penelitian ini adalah data kedatangan turis asing ke Indonesia. English : The data used in this study is the data of foreign tourist arrivals to Indonesia. Students consider that sentence 12 is the most readable, acquiring the highest score of 4,24. In other words, students can understand the sentence as observed that it is shorter than sentence 7, and the English translation conveys the same message as that in the SL. Meanwhile, the acceptability score reaches 4,12, which is also good. Thus, this sentence is considered to follow the TL rules quite well in terms of English grammar and rules. Next, Table 5 shows the acceptability and readability scores of Google Translate from the lecturers’ point of view. Table 5 Average Scores of Lecturers’ Perception on the Acceptability and Readability of Google Translate Sentence Number Average of Acceptability Score Average of Readability Score 1 3,33 3,33 2 3,67 3,67 3 2,67 3,33 4 3,00 3,33 5 3,67 3,67 6 3,00 3,33 7 2,00 2,00 8 2,00 2,67 9 3,00 3,00 10 2,00 2,00 11 3,00 3,00 12 3,33 3,33 13 2,67 3,00 14 3,67 3,67 15 3,33 3,33 16 2,33 3,00 17 2,33 3,00 18 2,00 3,00 19 1,67 1,67 20 2,67 3,00 Average 2,77 3,02 Max 3,67 3,67 Min 1,67 1,67 The highest score of acceptability and that of readability are exactly the same, 3,67, which happens to the same sentences; 2, 5, and 14. The lowest score also occurs in both criteria, 1,67, and also in the same sentence, 19. These results are different from those of students’ opinions in that the lecturers consider that Google Translate’s average level of acceptability is below 3 (2,77 out of 5) while the average level of readability is 3,02, which can be considered as moderately good. Example 3. This example refers to sentence 19 in Table 5. Indonesian : Aplikasi ini dapat membantu memudahkan pengguna Twitter yang menggunakannya sebagai media pemasaran atau promosi untuk melakukan promosi dengan cara melakukan tweet promosi terhadap follower yang sudah diklasifikasikan. English : This application can help make it easier for Twitter users who use it as a marketing or promotional media to do promotions by tweeting promotions to followers who have been classified. The lecturers consider that this sentence is not quite acceptable nor readable, shown by the average score of 1,67, which means below 2 (out of 5). It is observed that Google Translate has translated the Indonesian sentence quite literally in that it translates phrase by phrase. The sequence of the phrases is exactly the same as shown in the verb phrase “dapat membantu memudahkan pengguna Twitter” translated into “can help make it easier for Twitter users”. The suggested translation is “can help ease the Twitter users.” In addition, the sentence is possibly redundant as it repeats the same base word three times: promotional, promotion, and promotion. The explanation can be that the Indonesian sentence is written that way; thus, Google Translate is revealed to translate literally. Example 4. Another example of analysis is of sentence 2 from Table 5, which receives the highest score (with sentence 5 and 14) regarding both criteria. Indonesian : Ketujuh faktor tersebut adalah, kualitas website, kualitas citra produk, nama baik merek, pelayanan customer service, metode pembayaran, waktu, dan harga. English : The seven factors are, website quality, product image quality, brand name, customer service, payment methods, time, and price. It can be seen why sentence 2 receives the highest score. The most possible reason is that the sentence is simple and consists of a description of several factors, which is separated by a comma for each factor. However, ‘nama baik merek’ is translated into ‘brand name’, which does not include the lexical item of ‘baik’. In Indonesian, ‘nama baik’ is idiomatic; 213On Google Translate:.... (Menik Winiharti, et al.) thus, it should not be separated as ‘nama’ (name) and ‘baik’ (good) (Kamus Besar Bahasa Indonesia, n.d.). A suggested translation for this phrase is ‘reputation’; thus, ‘nama baik merek’ should be translated into ‘brand reputation’. Apart from such analysis, it is observed that punctuation also matters for sentence 2. In the SL, there is a comma after the verb ‘adalah’ (are), which is not supposed to be put there. The translation follows what the SL has written; Google Translate also puts a comma after the verb ‘are’. In other words, Google Translate still translates quite literally and does not pay attention to punctuation marks. Furthermore, Table 6 deals with the comparison between the average score of students’ and lecturers’ acceptability of Google Translate. Table 6 Acceptability Scores of Students’ and Lecturers’ Perceptions on Google Translate Sentence Number Students Lecturers 1 3,76 3,33 2 3,96 3,67 3 3,44 2,67 4 3,36 3,00 5 3,48 3,67 6 4,16 3,00 7 3,28 2,00 8 3,72 2,00 9 3,60 3,00 10 3,56 2,00 11 4,08 3,00 12 4,12 3,33 13 3,84 2,67 14 4,12 3,67 15 3,80 3,33 16 3,72 2,33 17 3,72 2,33 18 4,16 2,00 19 3,88 1,67 20 4,04 2,67 Average 3,79 2,77 Max 4,16 3,67 Min 3,28 1,67 Table 6 implies that students and lecturers have different perceptions regarding the acceptability level of Google Translate. On one side, students consider that Google Translate’s acceptability is slightly below 4, which means the translation is fairly acceptable (3,79 out of 5). However, lecturers’ average score on this criterion is below 2, which means that Google Translate is not acceptable enough (2,77 out of 5). It can also be shown from the comparison between the highest scores (4,16 versus 3,67) and the lowest scores (3,28 versus 1,67) of students and lecturers, respectively. Finally, the comparison between students’ and lecturers’ opinions deals with Google Translate’s readability. It is displayed in Table 7. Table 7 Readability Scores of Students’ and Lecturers’ Perceptions on Google Translate Sentence Number Students Lecturers 1 4,00 3,33 2 4,16 3,67 3 3,80 3,33 4 3,36 3,33 5 3,52 3,67 6 4,20 3,33 7 3,36 2,00 8 3,76 2,67 9 3,68 3,00 10 3,64 2,00 11 4,20 3,00 12 4,24 3,33 13 4,12 3,00 14 4,12 3,67 15 4,00 3,33 16 3,72 3,00 17 3,56 3,00 18 4,20 3,00 19 4,04 1,67 20 3,88 3,00 Average 3,88 3,02 Max 4,24 3,67 Min 3,36 1,67 Students’ average scores of Google Translate readability are higher than lecturers’, 3,88 and 3,02 respectively. Surprisingly, the range of the lowest scores is quite significant, 3,36 versus 1,67. The lecturers consider that sentence 19 is not comprehensible (1,67 out of 5), while students consider this sentence quite understandable (4,04 out of 5). The investigation on sentence 19 (Example 3) in the previous paragraphs is likely to explain why students and lecturers have significantly different opinions on this translation. Students still understand the sentence, but lecturers may possibly evaluate by looking at the unnatural arrangement of the phrases. This different perception is possibly due to the fact that undergraduate students and lecturers have a different background of knowledge. Even though their TOEFL scores are in the same range, their knowledge background regarding their academic field is different since the lecturers are master graduates. CONCLUSIONS The findings highlight several key points. First, more than half of the participants often use Google 214 LINGUA CULTURA, Vol. 15 No. 2, December 2021, 207-214 Translate, while the rest use it sometimes. It means that none of them has never used Google Translate web. In other words, the academics – undergraduate students and lecturers – are part of the users of Google Translate. Second, the students consider that the acceptability of Google Translate is quite good, while the lecturers think that the translation is rather good. Then about the readability of Google Translate, the students also view that the translation is quite understandable. Meanwhile, the lecturers’ point of view on this criterion is moderately understandable, even though such view is not evenly distributed to all sentences, which is shown by the lowest score of being below ‘rather good’. Thus, it can be said that to some extent, the students and lecturers have a different perception of Google Translate results. This is much likely because the students and lecturers have different knowledge backgrounds, then they assess the translation differently. In addition, apart from the research questions, it is indicated that Google Translate still translates literally and is not able to construct a varied or paraphrased structure as compared to that of the source language text. This can be an idea for further research which examines the strategy applied by Google Translate and the impact on its translation. Lastly, the research is not intended to advocate people, especially those having academic careers, to use Google Translate extensively. Rather, it is meant to provide a suggestion that Google Translate be much better used with post-editing stages, which is indeed done by humans. Thus, human translators or editors are still required to produce good translations. REFERENCES Aiken, M. (2019). An updated evaluation of Google Translate accuracy. Studies in Linguistics and Literature, 3(3), 253-260. http://dx.doi.org/10.22158/sll.v3n3p253. Alsalem, R. (2019). The effects of the use of Google Translate on translation students’ learning outcomes. Arab World English Journal for Translation and Literary Studies, 3(4), 46-60. https://dx.doi. org/10.2139/ssrn.3483771. Arsyad, S., Purwo, B. K., Sukamto, K. E., & Adnan, Z. (2019). Factors hindering Indonesian lecturers from publishing articles in reputable international journals. Journal on English as a Foreign Language, 9(1), 42-70. Brahmana, C. R. P. S., Sofyan, R., & Putri, D. M. (2020). Problems in the application of Google Translate as a learning media in translation. Language Literacy: Journal of Linguistics, Literature, and Language Teaching, 4(2), 384-389. https://doi.org/10.30743/ ll.v4i2.2893. Habeeb, L. S. (2019). Investigate the effectiveness of Google Translate among Iraqi students. Opcion, 35(21), 2899-2921. Hartono., Arjanggi, R., & Praptawati, D. (2019). Self- efficacy of Indonesian non-English lecturers in writing English academic papers for international publication. Advances in Social Science, Education and Humanities Research, 188, 28-31. https://dx.doi. org/10.2991/eltlt-18.2019.6. Hoi, H. T. (2020). Machine translation and its impact in our modern society. International Journal of Scientific and Technology Research, 9(2), 1918-1921. Karjo, C. H., & Metta, E. (2019). The translation of lexical collocations in undergraduate students’ theses’ abstract: Students versus Google Translate. Lingua Cultura, 13(4), 289-296. https://doi.org/10.21512/ lc.v13i4.6067. Kartika, D., & Priyatmojo, A. S. (2018). Analysis of Google Translate’s quality in employing translation techniques. Journal of English Language Teaching, 7(1), 40-49. Khosravizadeh, P., & Pashmforoosh, R. (2011). Google Translation: A semantic structure analysis. Retrieved from https://www.researchgate.net/ publication/216597782_Google_Translation_A_ Semantic_Structure_Analysis. Larson, M. L. (1998). Meaning-based translation: A guide to cross-language equivalence (2nd Ed.). Maryland, USA: University Press of America. Nababan, M., Nuraeni, A., & Sumardiono. (2012). Pengembangan model penilaian kualitas terjemahan. Kajian Linguistik dan Sastra, 24(1), 39-57. Setiyadi, D., Kuswardani, Y., Sari, D. K., & Martanti, D. A. (2020). Analyzing on English-Indonesian culture – Specific concept translation by Google Translate. International Journal of Scientific & Technology Research, 9(1), 2242-2246. Siregar, R., Risnawaty, R., Arfanti, Y., & Sembiring, M. (2020). Reflection of undergraduate students on translation process: An outlook of translation teaching in university. Lingua Cultura, 14(1), 57-67. https://doi.org/10.21512/lc.v14i1.6232. Valijärvi, R. L., & Tarsoly, E. (2019). Language students as critical users of Google Translate: Pitfalls and possibilities. Practitioner Research in Higher Education, 12(1), 61-74.