TRANSLATION ERROR USING GOOGLE TRANSLATE
Name: Naurah Nabilah Sya'bani Padmasehana
Abstract
During the
last decades, Google Translate, as a statistical machine translation (SMT), was
in the center of attention for supporting 90 languages. Although there are many
studies on Google Translate, few researches have considered other language –
English translation pairs. This study used Keshavarz’s (1999) model of error
analysis to carry put a comparison study between the raw English – Persian
translations and Persian – English translation from Google Translate. Based on
the criteria presented in the model, 100 systematically selected sentences from
an interpreter app called Motarjem Hamrah were translated Google Translate and
then evaluated and brought in different tables. Result of analyzing and
tabulating the frequencies of the errors together with conducting a chi-square
test showed no significant differences between the qualities of Google
Translate from other language to English and English to other language. In
addition, lexicosemantic and active/passive voice errors were the most and
least frequent errors, respectively. Directions for future research are
recognized in the paper for the improvements of the system.
KEYWORDS: Error Translation, Google Translation,
Machine Translate
Introduction
The use of
Google Translate has been increasing either in the academic discipline or in
the non – academic discipline. Despite the fast – turnaround time produced by
Machine Translation has been considered such as Google Translate, the quality
of the translation has been considered far from perfection. Thus in order to
evaluate the quality of machine translation, error analysis has been suggested
to be conducted. In line with this, numerous text genres have been
investigated, including one of which is an abstract text. Abstracts as a
summary of a research paper harbors important information where it serve to
attract readers to whether read the entire passage or leave it.
In Indonesia, the need to translate abstracts
of undergraduate thesis into English is a requirement for students to complete
their final year academic papers. Some universities oblige students to have the
abstract of their scientific translated by official institutional language
center, but, some others do not require official translation service to
translate their abstract into English. As a result, students are allowed to use
whatever means available to translate their abstract as fast as possible
without taking into account the accuracy of the translated text. The last
resort students would take is using Google Translate to translate their
abstract since the service it provides is free and real – time basis results
within seconds. However, the majority of
students have been unaware of the consequences it bears upon it such as wrong
world choices, wrong prepositions, wrong word order and so on. They have been
known to have used Google Translate to translate their abstracts of bachelor’s
papers into English. This phenomenon is found in the Methodist University of
Indonesia, Medan. The university administrator does not require final year
student to translate their bachelor’s papers’ abstracts accurately using the
service of professional translator.
International Journal of English Language and
Linguistics Research
Vol. 5, No. 2, pp. 15-23, April 2017
Published by European Centre for Research Training and
Development UK (www.eajournals.org)
Background of Study
Nowadays,
English is considered as international language due to the impact of
globalization. In the process of understanding English, translation is needed.
According to Catford (1965:20), translation is the replacement of textual material
in one language (SL) by equivalent textual material in another language (TL).
Its uses can be found in the translation of textbook, state documents, literary
works, bilingual books, business documents, journals or scientific works and so
forth. Hence, because of its vital role, translation can offer a solution to
overcome language gap in communication.
Indonesian
people often use internet in their daily life. One of the tools on the internet
that can help them in translation process is Google Translate service. Google
Translate is a convenient tool that offers free instant translation service on
the web. It can be utilized to translate words, clauses, sentences, paragraphs
and even a web page between any pairs of supported languages. Moreover it can
be utilized to minimized time and effort to do translation tasks because the
translation results are instantly generated. The translation is also helped
with the easiness and availability of Google Translate. Which are online and
accessible to anyone and anytime for free with internet connection.
Meanwhile,
the use of Google Translate to translate has brought some issues. Some
translators might use Google Translate blatantly without any revising effort on
its translation, which leads to some overwhelming translation results. Google
Translate itself also has limitation, that when translating complex sentences,
it would sometimes resulted in inaccurate translation. There is an opinion that
using Google Translate to do the translation work is too narrow – minded and
easy because anyone can copy the text to Google Translate, choose the language
and press the translate button. Even, anyone who does not have any proficiency
in both source and target language is able to do it. Despite of those issues,
the use of Google Translate to do the translation work is indeed fast and can
bring an instant overview of translation result.
From those
issues, there comes an idea of analyzing the errors in translation result of
certain type of text by using Google Translate. One type of the text, which is
very familiar in daily life, is News Item text. News Item texts can be found in
newspaper, magazines, blogs, news, websites and so forth. Utilizing Google
Translate to translate News Item texts is also popular due to the need of fast
information updates in the society in a form of translating foreign news into
local language that can be understood by local people and vice versa. Thus,
this research was employed to find and explain errors in Google Translate’s
translation results of News Item texts from Indonesia to English.
Research Finding
Translation
Translating is the written transfer activity text
messages from one language texts (Hoed; 2006: 5). In this case, the translated
text is called the source text and a language called the source language. With
regard to the translation, the text drafted by the translator is called the
target text and the language is called the target language. Nida and Taber
(1994:34) stated “translation is the reproducing message in the source language
with natural equivalence in the target language, play through two steps, first,
based on the meaning and second based on style”. In other words, the translation
is to reproduce the message in the source language with their natural
equivalents in the target language, in two steps, first, based on the meanings
and the second based on the style (the language) it.
Based on some of the above definitions of translation,
translation is a section on the relationship between two or more languages that
then either transfers the meaning of the source language (SL) into target
language (TL).
Types of
Translation
Translation
can be classified into several types. According to Lado (1968; 261 – 262)
translation is not only used for formal purposes, but the translation is also
used for informal purposes. Therefore, the translation may occur in two areas,
namely in the realm of factual formal translation in literature (literary) in
the realm of informal translation.
Larson
(1928: 35) stated that translations into shapes (form – based information) and
translation of the meaning – based translation). Examples of forms – based
translation are a literal translation, while translating idiomatic translation
is an example of the meaning – based translation. Literal translation is the
translation of the word per word (word to word translation). This translation
is usually used in linguistic translation. In idiomatic translation (idiomatic
translation), the translator tries to divert the meaning of sources language
into target language so that can be understood easily and naturally. Therefore,
the emphasis on the idiomatic translation of the meaning or message is not on
words or other lexical items.
Moreover,
Newmark (1988: 45 – 47) classifies translation into eight kinds, namely:
1. Word for Word Translation
In
translating word for word translation, the wording (word – order) sources
languages is maintained and words in sources language translated one by one in
accordance with the general meaning and not taking account of the text.
2.
Literal
Translation
In this
translation, grammatical constructions source (source language) was transferred
into the target language grammatical constructions (target language) closest,
but the words translated lexical still single, out of context.
3.
Faithful
Translation
In this
type of translation, contextual meaning diverted from source language into
target language, despite the limitations of target language grammatical
structure. Words cultures are transferred and the degree of “abnormality”
grammatical and lexical persists.
4.
Semantic
Translation
This
translation is promoting the values the beauty of sources language. Translating
this model more flexible by providing a space for creativity and intuition
interpreter.
5.
Adaptation
Translation
This type
of translation is a form of translation ‘most free’, commonly used in drama and
poetry.
6.
Free Translation
In this
type of translation, the message and mandate reproduced, regardless of the form
in the source language. In other words, in this kind of translation, ‘content’
is translated without following the ‘shape’ as in source language.
7.
Idiomatic
Translation
In
translation this type of message or mandate reproduced in target language but
there is a tendency to distortion shades of meaning, due to the use of idiom
that was not there at source language.
8.
Communicative
Translation
In this
type of translation, contextual meaning source language can be received and
understood by the target audience of the translation.
Machine
Translation
Machine
translation is a blend of linguistics with computational science, often
referred to as part of the science of computational linguistics. The point of
this science tries to make the machine capable of translating one language into
another. Besides proving that the translation is no longer solely a translation
by humans, but in a professional context, an improvement of process and
products that combine the power of computer and the analysis of language –
based computers with the human ability to analyze the meaning and determine an
appropriate form into other languages.
Google
Translate is one of several machine translations most commonly used by people
around the world to translate text over 90 different languages. Google
Translate searches different documentaries to find the best appropriate
translation pattern between translated texts by human. This pattern searching
is called Statistical Machine Translation. Since the number of translated texts
varies from users to users, consequently, the quality of Google Translate
depends on the number of human translated texts searched by Google Translate
(Karami, 2014). Quite recently, another assessment to the study of Google
Translate does not handle subject – verb agreement very well while translating
English sentences into another language compare to human translation.
Error Analysis
Error Analysis
Error
analysis (EA), a fundamental branch of applied linguistics, emerged in the
sixties to address students’ performance (Shrestha, 1979). According to Longman
Dictionary of Language Teaching & Applied Linguistics (2010), (EA) is
manifested in order to (1) describe strategies used by the learners in language
teaching, (2) spot causes of errors and finally (3) gain information on common
difficulties in language learning to develop materials and strategies to help the learners
avoid their errors.
Error
analysis is a type of linguistic analysis that focuses on the errors learners
make. It consists of a comparison between the errors made in the Target
Language (TL) and that TL itself (Corder, 1974). According to Richards et al.
(1992: 96), “error analysis may be carried out in order to: a) find out how
well someone knows the language, b) find out how a person learns language and
c) obtain information on common difficulties in language learning”. Moreover,
error analysis helps to identifying the weaknesses, with a variety of
techniques, for identifying, classifying and systematically interpreting the language
learners’ errors (Khodabandeh, 2007).
In terms of
errors made by student of English, errors appear when the learner’s knowledge
of the rules of the target language is incomplete. Errors is considered to be
systematic, governed by the rule and also regarded as rule – governed when they
follow the rules of the learner’s inter language (Keshavarz, 2011). According
to Abbasi and Karimnia (2011) it is essential that teachers be able to adjust
their teaching plan to make their teaching work more effectively by identifying
learners’ errors. Moreover, recognizing errors can provide valuable information
for teachers about how much the learners has learned and what kind of problems
s/he has in the study of language. As Conde (2011) puts it, “error detection has
been the traditional basis for translation evaluation”. Gass and Selinker
(1994) identified six steps followed in conducting an error analysis. These
included ‘collecting data’, ‘identifying errors’, ‘classifying errors’, ’quantifying
errors’, ‘analyzing sources of error’ and ‘remediating for errors’.
Error in Translation
According
to Kohler (1979: 216), based on the concept of equivalence between source text
and target text, a translation error is regarded as some kind of non –
equivalence between source text and target text or non – adequacy of the target
text. In functionalistic approach and approaches based on the ‘skopos theory’,
an error is defined as relative to the fulfillment of the target text function
and the receiver’s expectations (Schmitt 1998:394; Nord 2009; 190).
Conclusion
Conclusion
Using
Google Translate or any kind of Machine Translator is not right an answer to
translate for translating foreign language. Google Translate is only use to a
pre – translation that still needs to be revised in linguistic aspects so it
will not cause mistranslation in the future for someone who needs to translate
foreign language fast.
References
Napitupulu, Sependi (2017) “ANALYZING INDONESIAN –
ENGLISH ABSTRACTS TRANSLATION IN VIEW OF TRANSLATION ERRORS BY GOOGLE TRANSLATE”.
From http://www.eajournals.org/wp-content/uploads/Analyzing-Indonesian-English-Abstracts-Translation-in-view-of-Translation-Errors-by-Google-Translate.pdf&ved=2ahUKEwj5xLGV7-nbAhXZMt4KHTCgDk4QFjABegQlAhAB&usg=AOvVaw0RkTg3D9hEr-Qa08UT4WAD
accessed on June 19th 2018
Ismail, Adam and Hartono Rudi (2016) “ERRORS MADE IN
GOOGLE TRANSLATE IN THE INDONESIAN TO ENGLISH TRANSLATIONS OF NEWS ITEM TEXTS”.
From http://www.journal.unnes.ac.id/sju/index.php/elt/article/view/11228 accessed on
June 19th 2018 14:54
Anggaria, Aria Septi and Muhamad Sofian Hadi (2017) “LINGUISTIC
ERRORS ON NARRATIVE TEXT TRANSLATION USING GOOGLE TRANSLATE” From
http://www.academia.edu/36368688/LINGUISTIC_ERRORS_ON_NARRATIVE_TEXT_TRANSLATION_USING_GOOGLE_TRANSLATE accessed on June 19th 2018
Stankeviciute, Gilvile, Ramune Kasperaviciene and
Jolita Horbacauskiene (2017) “ISSUES IN MACHINE TRANSLATION”. From http://www.content.sciendo.com/view/journals/llce/4/1/article-p75.xml
accessed on June 19th 2018
Koponen, Maarit and Leena Salmi (2015) “ON THE
CORRECTNESS OF MACHINE TRANSLATION: A MACHINE TRANSLATION POST – EDITING TASK”.
From http://www.jostrans.org/issues23/art_koponen.pdf accessed on June 19th 2018 21:32
Ghasemi, Hadis and Mahmood Hashemian (2016) “A
COMPARATIVE STUDY OF GOOGLE TRANSLATE TRANSLATION: AN ERROR ANALYSIS OF ENGLISH
– TO – PERSIAN AND PERSIAN – TO – ENGLISH TRANSLATION” From http://dx.doi.org/10.5539/elt.v9n3p13
accessed on June 19th 13:33
Afshin, Hakiminejad and Mohammad Ali Alaeddini (2016) “A
CONTRANSITIVE ANALYSIS OF MACHINE TRASLATION (GOOGLE TRANSLATE) AND HUMAN
TRANSLATION: EFFICACY IN TRANSLATING VERB TENSE FROM ENGLISH TO PERSIAN”. From
http://www.mcser.org/journal/index.php/mjss/article/viewFile/9502/9172 accessed on June 19th 2018 22:01