THE EFFECTIVENESS OF E-LITRANS TOOL FOR TRANSLATING FIGURES OF SPEECH FROM INDONESIAN INTO ENGLISH

Difficulties in translating figures of speech should not be allowed to occur for a long time. There must be an effort to solve the problem immediately. Introducing electronic devices to facilitate the translation of figures of speech in the form of a trans tool is an important effort to do. One of the efforts is to use the E-Litrans Tool as an electronic device that can improve the quality of the figures of speech translation. This trans tool was tested in experimental research of translating figures of speech from Indonesian to English and vice versa. Three figures of speech, metaphor, simile, and personification, were focused to analyze in this research because they are more widely used in everyday life in the Indonesian context. The three types of figures of speech are arranged into a set of test items, each consisting of 5 pieces so that there are 15 figures of speech. One set of test items was given to the experimental group and the control group. The experimental group consisted of 22 students while the control group consisted of 25 students. All of these students came from translation classes that we taught. The experimental group was treated using the E-Litrans Tool in translating figures of speech, while the control group was not. Based on the research results, it was known that the E-Litrans Tool was able to significantly improve the translation quality of the experimental group. The average translation quality score of the experimental group after using the E-Litrans Tool increased from 60.2 to 95.8. It means that there was an increase in the average value of 35.6. the overall mean score of their translation quality is 78.36. The translation quality for the control group that did not use the E-Litrans Tool was insignificant and dropped drastically. The mean pretest and posttest scores were 56.88 and 55.84 or the overall mean score of their translation quality was 56.4 without the progress, even dropped decreasingly up to 1.04. It can be concluded that E-Litrans Tool increased the figures of speech translation quality scores significantly. Therefore, we suggest that translators can use the E-Litrans Tool to improve the quality of their figures of speech translation results. The followings are examples of translation naturalness found in the students’ translation products from the pretest and posttest of the experimental group that applied the E-Litrans Tool (E-LT). The source texts (ST) were translated, transferred, or replaced to the target text (TT) naturally.


Introduction
Translating figures of speech in terms of metaphor, simile, and personification from Indonesian into English is a challenging and complicated translation process (Issa, 2017). This can be proven by a number of translations that are not natural. For example, the Indonesian simile 'seputih kapas' is often literally translated into English simile 'as white as cotton' whereas in English culturally there is no such simile. In this case, English has a similar figurative equivalent, namely 'as white as snow'. It is very difficult for translators to translate it if they do not have a large cultural knowledge of both languages. If the simile is translated literally, it will seem unnatural (Hartono & Priyatmojo, 2015). The other example is the translation of metaphor "My love is a red, a red rose'. This metaphor should be translated into the Indonesian simile 'Kekasihku bagaikan sekuntum mawar merah' in order to be natural, not to be translated into 'Cintaku adalah sekuntum bunga mawar merah' that seems literal. Next, what happened to the translation of the following personification? The Indonesian personification 'Angin menangis di gelap malam' is usually translated literally into English personification 'The wind cried in the dark' although if is translated back to Indonesian, the meaning will be 'Angin menangis di kegelapan' because literally the meaning of 'the dark' is similar to 'kegelapan' not 'gelap malam'. Thus, the personification translation is not natural and will not be accepted in the English culture.
In connection with the problem of translating figures of speech, many researchers in the field of translation commented on this problem. One of them is Teilanyo (2007) who says that translating figurative language or figure of speech is a sensitive task. The translators must be able to maintain the content of the source text message which is firmly embedded in the text, so that in translating it into the target language, the translator must be able to find the most effective methods and techniques. The message in this figure of speech must be translated faithfully, but the translation must be natural, easily understood by the target language user and acceptable in the cultural context of the recipient. If the figure of speech is difficult to translate, the translator can provide annotation to explain the figure of speech he translates. The footnote technique might be an alternative for translating figures of speech that have no equivalent in the target language. Ayomi (2009) adds that translating figures of speech is a challenging task. Translators must find the same meaning in various forms in the target language. Several strategies can be carried out by the translator to maintain the meaning and aesthetic elements of the source text, which must be adapted to the conditions of the text, namely translating directly if possible and the results are natural in the target language, changing metaphors into similes, replacing with a language style that has the same meaning in target language, translate directly with explanations and translate the meaning explicitly without using figurative language.
From the cases above, we can infer that translating the figures of speech seem not only translating forms of language but also styles (Isaa, 2017). The translators have difficulties in determining the nature and the closest equivalent of metaphor, simile, and personification in the target language. It means that they have primary problems with naturalness because literary texts or stylistic words contain connotative meaning that is difficult for translators to translate (Lestiyanawati, 2014). Translating literary words or figurative languages require translators' deep understandings. If translators do not understand the meaning, their translation products can be unnatural. Translating literary words is similar to translating the Culturally Specific Items (CSIs). Here translators must be able to search the closest equivalent words (Bagheridoust, 2017). The CSIs always have unique meanings that must be transferred into target languages naturally. This uniqueness exists implicitly in the beautiful words like metaphor, simile, and personification, and other types of figures of speech. The literary words have translation difficulty levels, even they are untranslatable.
To respond to the above issue of literary translation, especially on the translation quality of naturalness, the E-Litrans tool comes as a solution. This tool is an online translation application designed for translating proverbs, idioms, and figures of speech. It has potential properties as a translation tool to translate beautiful words naturally and acceptably. Principally this translation device is like an online electronic literary translation tool or translation memory that is specifically designed to help translators translate idioms, proverbs, and figures of speech from English into Indonesian and from Indonesian into English (Hartono, Sakhiyya & Priyatmojo, 2019). E-Litrans tool is designed to translate Indonesian proverbs, idioms, and figures of speech idiomatically to English proverbs, idioms, and figures of speech or vice versa. It uses free, adaptation, idiomatic, and communicative methods. The tool is packaged with techniques of addition, deletion, amplification, modulation, and description. This translation tool has many words, phrases, sentences in a bilingual system (English-Indonesian and Indonesian-English). Not only does it serve for searching texts, but it is also designed to load other bilingual proverbs, idioms, and figures of speech as new entries inserted or contributed by users, as far as they are really new entries, or not the already available input in the E-Litrans Tool. Contributors can insert or add new entries by using a username and password given by the admin. After the new entries added, the admin will censor and validate them first whether they are available or not in the memory system. If it is not available and really new, the entries will be validated and listed in this software (Hartono, Sakhiyya & Priyatmojo, 2019 After the steps are taken, the new entries will be saved, checked, and validated by the admin. The new entries will be accepted and stored if they are new and appropriate with the system. Till nowadays the E-Litrans Tool has more than 500 bilingual entries of literary texts (Hartono, Sakhiyya & Priyatmojo, 2019.

Method
To see the effectiveness of the E-Litrans Tool (E-LT), we conducted True-experimental research by involving 22 students for the experimental group (EG) and 25 students for the control group (CG). Hence, there were 47 students as the research participants chosen totally by using the total sampling technique (Etikan, Musa & Alkassim, 2016). The EG was given the treatment by using (E-LT) while the CG was not. The CG was only asked to use other electronic dictionaries and machine translations, such as Google Translate and U-Dictionary to translate their translation tasks.
The data source used in this research was a set of test items that consist of 5 Indonesian proverbs, 5 Indonesian metaphors, 5 Indonesian similes, and 5 Indonesian personifications. The test items were set based on the final test designed that had been validated and tried to be tested before in the preliminary research. The test given was the translation test from Indonesian into English. The EG used E-LT to translate the texts while CG used other translation tools.
The test items were given to both the experimental group and the control group. They were assigned to do their task for 100 minutes. After the test finished, the test results were collected and grouped based on the experimental and control group. After the test results were collected and classified based on the group, we classified them into a proverb, metaphor, simile, and personification. Then, we analyzed the translation result group by group and classification by classification. Each test item was assessed by comparing the Indonesian texts as the source language to the English translated texts as the target language. Each translation was analyzed by using the Contrastive Analysis model (Presada & Badea, 2014). Through this way, we could see the quality of their translation results.
In this research we used the scale of naturalness level to assess the translation quality (See table 1). The scale is usually used to assess the proverbs, idioms, and figure of speech translation from English into Indonesian and vice versa (Hartono, 2018 Translation feels natural; the term used is common and familiar to the reader; words, phrases, clauses, and sentences used are in accordance with the rules of Indonesian language. Natural 2 (34-67) In general, the translation already feels natural, but there are few problems with the use of the term; there was a slight grammatical error.
Less natural 1 (0-33) Translation is not natural or feels awkward and the term used is unusual; words, phrases, clauses, and sentences used are not in accordance with the rules of Indonesian language.
Not natural The scoring system of the test result was firstly taken from scale 100 that then was converted to the scale 3. The score conversion is scale 3 = 68-100, scale 2 = 34-67, and scale 1 = 0-33. In this research, we only focused on the naturalness level of translation. In analysing this translation level quality, we use the scale of naturalness level (See table 1). Each test item or the translation result was scored by using scale 3 for the translation which is natural, scale 2 is for less natural, and scale 1 is for not natural. The scoring results then were calculated statistically by using t-Test formulation. Data management and analysis were performed using SPSS v22.

Findings and Discussion
This part provides the findings and discussion based on the pretest and posttest results from the experimental group (EG) using the Android-based E-Litrans Tool and the control group (CG) using other translation tools (e.g. Google Translate and U-Dictionary). Table 3.1 shows the experimental data on the experimental group that used E-Litrans Tool (E-LT) while table 3.2 presents an overview of the pretest and posttest results of the control group that did not use E-LT. The scores shown in tables 3.1 and 3.2 are still in the 0-100 range, they have not been converted to numbers in the range 1-3.  Table 3.1 shows that the experimental group achieved a significant improvement in their translation products after using the E-Litrans Tool Application in the translation process. It can be seen from the average mean of both the pretest and posttest results. The average mean score of the pretest is 60.2 and the average mean score of the posttest is 95.8. If it is averaged again across these averages, the overall mean score of their translation quality is 78.36 with the progress score up to 35.6. It is a very excellent improvement gained by that group after using the E-LT in their translation process. From the data, by referring to the translation rating scale based on the level of naturalness, it can be said that the translation results produced by the experimental group are natural. This is indicated by the average pretest-posttest translation score of 78.36. It illustrates that the score is in the position of the 3-scale value, which means that the translation result is natural.  Table 3.2 presents the breakdown of experimental data of the control group that achieved a very low improvement of their translation products. It can be seen from the average mean of both the pretest and posttest results. The average mean score of the pretest is 56.88 and the average mean score of the posttest is 55.84. The overall mean score of their translation quality is only 56.4 without the progress, even dropped decreasingly up to 1.04. It was because they only used Google Translate and other translation tools or apps to translate their translation tasks. From the data above, by referring to the translation rating scale based on the level of naturalness, it can show that the translation results produced by the control group are less natural. This is shown by the average pretest-posttest translation score of 56.4 describing that the score is in the position of the 2-scale value, which means that the translation result is less natural.
Apart from the table, we can also see an increase of the translation score from the chart. The bar charts below show the progress scores of both experimental group and control group in which the student got increased the score before the pretest and post-test can be seen clearly.
As can be seen from the chart 3.1 (below), the experimental group reported significantly more increasing scores of translations than the control group. All students' translation scores from pretest to Language, Education, and Development (LEAD) Journal ISSN: 2827-976X Vol 2, No. 4 (2021) 28 -36 http://jos.unsoed.ac.id/index.php/lead/index posttest increased even though the difference in increase was different. This proves that the E-Litrans Tool is able to improve their translation quality to achieve natural translation results.

Chart 3.1 The Pretest and Posttest Results of EG with E-LT Application Chart 3.2 The Pretest and Posttest Results of CG without E-LT Application
On the other hand, chart 3.2 shows that the progress of the translation score before and after the test shows various increases, stagnation, and even decreases in scores. In fact, the chart depicts more of the stagnation and decline in the translation score even though based on the scale of the score in table 3.2 it reaches a value of 2 (less natural). There were only five students who experienced an increase in their scores, namely students' number 5, 13, 17, 21, and 24, while the rest did not change and even decreased. There were seven students whose translation scores were stagnant, not changing at all, namely students' number 1, 10,11,16,20,23, and 25 while students whose translation scores had decreased were those with numbers 2, 3, 7, 8, 9, 12, 14, 15, 18, 19, and 22 students have increased their translation scores, 32% of students whose scores have stagnated, and 45% of students whose translation scores have decreased. This proves that translating literary texts, especially metaphors, similes, and personifications cannot use Google Translate or an electronic dictionary. Machine translation and other applications besides the Android-based E-Litrans Tool cannot translate the literary text naturally.
On the other hand, it can be seen that the standard deviation of EG is 5.413 and the standard error is 1.133 while the standard deviation of CG is 9.600 and the standard error is 1.920. It means that the alternative hypothesis (H1) is accepted that states there is a significant difference in translation results between the EG that used E-LT and the CG that did not use E-LT (See table 3.3). Furthermore, to prove whether the above differences are significant or not, it is necessary to interpret them in the following Independent Samples Test (See table 3.4). Based on the output above, the Sig. Levene's Test for Equality of Variance is 0.053> 0.05, it can be interpreted that the data variance between the experimental group and the control group is homogeneous or the same so that the interpretation of the Independent Sample Test output table above is guided by the values contained in the Equal Variance assumed table.
Based on the Equal Variance assumed the Sig. (2-tailed) of 0.000 <0.05, it can be concluded that Ho is rejected and H1 is accepted. Thus, it can be concluded that there are significant differences between the two groups.
To clarify the values above, a comparison between the tscore and the t-table is made. It is known that the tscore value based on the calculation above is 9.516. The df value used is 45 matched at the value of 0.025 in the table which results in a ttable value of 2.01410. This shows the tscore > ttable, so it is confirmed that there is a significant difference between the experimental group and the control group in translating literary texts. See the distribution percentage point in the table 3.5 below.

TT: Keeping horns of a dilemma
Between the devil and the great sea.

TT:
The leaves are waving. The leaves waved in the wind.
Data 3, in the pretest column, we see that the source text of metaphor 'Mulutmu harimaumu' translated into the target text of metaphor 'Your mouth is your enemy' is not natural because the target metaphor is literal translation that is not available in the English culture as an equivalent metaphor. As for the posttest column, the Indonesian metaphor is translated naturally into the English metaphor 'Your tongue is fire'. Translation of this metaphor is natural and naturally acceptable in English culture (Farahani and Ghasemi, 2012).
According to data 4, the Indonesian simile 'in the pretest is not translated into the same simile. 'Keeping horns of a dilemma' is not the English metaphor, it is just a meaning of the simile not the English simile. The appropriate, natural, and closest English simile is 'Between the devil and the great sea'. This simile is in accordance with the original context, culture, and society of English. Thus, simile should be translated into simile (Moon, 2008).
In data 5, it can be seen that the personification "Leaves are waving" in the pretest column is not a natural personification, it is a literal translation of the personification 'Dedaunan melambai-lambai'. The appropriate and accepted personification in English, freely using addition technique, is 'The leaves waved in the wind', in the target text of the posttest, because it is accepted naturally in English. The translator added 'in the wind' to show the atmosphere that the leaves are blown by the wind, so that the atmosphere is more interesting and gives a literary impression.

Conclusion
The present study was designed to determine the effect of E-Litrans Tool (E-LT) on translating figures of speech, in this case they are metaphor, simile, and personification. The study has identified that E-Litrans Tool increased significantly the score of translation and improved the translation quality of naturalness. One of the most significant findings to emerge from this study is that the post-test results of the experimental group increased very high and all the results of the translation became natural, while the post-test results of the control group increased very slightly, stagnated, and even dropped dramatically. The result of the study indicate that E-Litrans Tool is an effective transtool that can improve the translation quality of metaphor, simile, and personification. The findings from this study make several contributions to current issues of translation tools for translating figure of speech. A limitation of the study is that the figure of speech studied is only a metaphor, simile, and personification, not including all figures of speech because the E-Litrans Tool Application itself is only designed to translate the three figures of speech. In spite of its limitations, the study certainly adds to our understanding of the figure of speech translation rules. It means that metaphor must be translated to metaphor, simile to simile, and personification to personification naturally from the source language to the target language. More information on X would help us to establish a greater degree of naturalness on the figure of speech translation. The findings of this study have a number of practical implications of using E-Litrans Tool for translating figure of speech from Indonesian into English and vice versa.
Ensuring appropriate system, services and support for E-Litrans Tool should be a priority for translation class students, teachers, and literary text translators.