Error Analysis of Role-Play Scripts Translated from Malay to Arabic Language Via Google Translate

Today, translation of one language into another not only involves humans but it can also be done through translation applications available on the internet. At first the ability of these apps was to translate one word but lately there have been applications that seem to translate not only words but can translate complete sentence and book phrases. The applications that are commonly used and well received is Google Translate application in which there is a translation of the Malay language to Arabic and vice versa. But how accurate is Google Translate to be used to translate the texts of the Malay language into Arabic. Therefore, this study aims to explore the accuracy of Google Translate in translating text from Malay to Arabic language and explain the errors found by the script role play provided by UiTM Melaka Branch students as part of an ongoing test during the semester. This study is expected to benefit Arabic language learners to minimize the use of Google Translate and to fine-tune the translation results especially in terms of vocabulary and sentence structure.


Introduction
Translating is an act of interpreting the meaning of a text and reproduce it in a different language. Translating is an important activity to disseminate knowledge because it enables the society to enjoy the knowledge that comes from foreign cultures. According to Kirsten (2012), translation is an activity that aims at conveying meaning or meanings of a given linguistic discourse from one language to another. Translation can be defined in terms of sameness of meaning across languages. Amirah (2017) also considered translation as a mental activity in which a meaning of given linguistic discourse is rendered from one language to another. It is the act of transferring the linguistic entities from one language in to their equivalents in to another language.
On the other hand, Nida and Taber (1982) stated that translating consists in reproducing in the receptor language the closest natural equivalent of the source language message. This definition is more comprehensive which explicitly state that translation is closely related to the problems of languages, meaning, and equivalence. In their book 'The Theory and Practice' mentioned that a translator must be searching for the closest match semantically to the original text in order to achieve the actual meaning in two different languages. A high competency in two different languages is crucial for a translator in order for the messages and ideas contained in a text can be conveyed efficiently to the target readers.
Normally, translating is a human task, however with the advancement of technology, translation work is now assisted by machine in network and it is used widely. Computerassisted translation can work automatically and faster compared to manual translation (Imran, 2003). Therefore, translation work done by machine does not require reference to any dictionary or thesaurus.
However, Iman (2010) stated that up to this date, translating using machine still faced with various weaknesses and the translations produced by the machine, just like Google Translate, consist of limitations and have not been producing perfect translation work. The quality of translation from one language to another is not always the same. Novia (2012) has the same opinion when she mentioned that the translation work by Google Translate is full of confusions in terms of grammar, choice of words as well as spelling.
Therefore, the researchers are interested in analysing the role play scripts produced by students who are taking the Arabic language subject TAC101 and TAC151 at Universiti Teknologi MARA Melaka Branch. Most students disregard the grammaticality of Arabic language when they write the scripts because they utilise normally machine translation such as Google Translate by translating Malay language scripts into Arabic language.

Literature Review
Undoubtedly, the translation field has benefited greatly from the advancement of technology today and the field itself has made full use of the development of technology. It helps us to keep up with the globalisation era where the access to information has become faster and more efficient. A translator also does not need to depend on reference collection and printed dictionary because with the use of electronic tools, the translation work has become faster and easier. In this twenty-first century, all translators make use of the internet connection, computer and other electronic tools in order to help their translation work (Khadijah & Maheram, 2013).
Nevertheless, the translation work done by Google Translate still needs to be reviewed and rechecked. Among the problems from Google Translate is the quality of grammar, therefore, the translation work produced is not up to the standard in terms of grammaticality (Mohd Khairulanuar, 2009). Another aspect that requires attention is the meaning accuracy and semantics (Akmar & Harun, 2016). Translation work from Google Translate requires refinement and editing work. Undeniably Google Translate has the capability to translate a text, however, the skills of a human translator or editor is still needed to produce a polished translation work. This is because, a human translator translates based on knowledge, skills and experience, while the work from Google Translate is limited to corpus and word bank (Khadijah & Maheram, 2013;Faezah & Hasnah, 2015).
In a research by Radiah & Eliza (2007) titled "Analysis of Matching Meanings in Computer-Assisted Translation", the researchers were able to provide evidence that using a computer-assisted program did not hinder a target text to correspond to its source text. The scope of the study, however, was limited to denotative translation of proper and common noun phrases only. Both of these aspects are what distinguish the current paper. To sum up, firstly, researchers admitted to the fact that they had to make quite a number of adjustments to the target text in order for it to be consistent with the source text as well as the cultural aspects of both the target and the source text. Secondly, computer-assisted translation software should be equipped with a more expansive and up-to-date corpus of analogous bilingualism. Thirdly, there arises a need for a statistical mechanism that can be used to efficiently find the translation of a word without disregarding its context, especially for a phrase that may have several distinctive definitions.
Amin & Naimah (2011) explored the topic of text translations from Arabic to Malay language from the viewpoint of semantics, whereas the current paper focuses on the reverse, which is the translation of texts from Malay to Arabic language. Such is the direction taken as the research instruments of this paper will be the texts written by diploma students of Hotel Management as well as Business Management who were enrolled in Arabic classes. Amongst the assignments which were required of the students were writing a role-play script. It was largely discovered that students would employ the use of Google Translate to assist them in the assignment.

Research Objectives
The research objectives are; 1. To explore the accuracy of Google Translate in translating text from Malay to Arabic Language. 2. To explain the errors found by the script role play provided by UiTM Melaka Branch students.

Methodology
The methodology used for this study is qualitative in nature, because the main concern of this study is to analyse script role play provided by 15 groups from Faculty of Hotel and Tourism Management and Faculty of Business Management, UiTM Melaka Branch. These students took Arabic language course as the third language. In this research, Google Translate is used as a medium to translate the script from Malay to Arabic language.

Results and Discussion
There are several errors especially in terms of vocabulary and sentence structure that need to be improved and refined.

Vocabulary Errors
The analysis of vocabulary errors can be seen in table 1 and 2.

Sentence Structure Errors
The analysis of sentence structure errors can be seen in tables 3, 4, 5 and 6.

3.
Itu buah-buahan import (Those are imported fruits) Amongst the errors found in the Arabic scripts written by the students are certain words in the source text were omitted in the target text and this resulted in translation inaccuracy. The same also happened to situations where students unknowingly employed inapposite words in the target language to refer to their source language.
Furthermore, there were also syntactic inaccuracies found in the Arabic translations written by students and amongst the issues were that of idhafah, pronouns or dhamir without distinguishing muzakkar and mu'annath, determiners or ism al-isyarah which again, are not discrete between muzakkar and mu'annath. The distinction between muzakkar and mu'annath again posed an issue for students as they failed to recognise this aspect of the target language in translating the verbs from their source language and this error sits atop as the preponderance of errors in target language. Most of these errors were the result of students being too dependent on translation software or applications such as Google Translate.
Nevertheless, despite all the shortcomings presented, translation software and applications are without a doubt essential in rendering a complicated work like language translation less demanding. Accuracy is possible but it can only be obtained given the knowledge, aptitudes, and experience of the translator. As useful as it is, Google Translate still functions best if it is not given the driver's seat in mediating the meaning between the source language and the target language. In other words, without basic knowledge in language translation, a translated text will still be at risk of being substandard and misleading to its intended audience despite it having been translated by a translation software of application.