Machine Translation: Source-Oriented and Target-Oriented Approaches

Abstract

Machine translation has become one of the important features in social media especially on Instagram. This case study investigates the English-Indonesian translations of sentences as the caption on Instagram, as well as analyses the source-oriented and target-oriented translations shown by the target language structures and styles. The frequency of occurrence is also calculated in order to analyze the different parts of sentences that are translated by applying source-oriented and target-oriented approaches. The data were collected from an Instagram account that has a long caption consisting twenty-nine sentences and also one phrase in English and their translations in Indonesian. The data were classified into the language expressions types and then analyzed based on the source-oriented and target-oriented translation approaches (Toury, 1980, 2012). The results show that there are three types of language expressions, they are common expressions, idiom, and phrasal verbs. In terms of the frequency of occurrence, 62% of all parts of sentences are translated based on the target-oriented approach. It reveals that the TL is good enough for the target readers to understand the messages and meanings of the original language. The common expressions are the most frequent parts of sentences that are translated based on the target-oriented approach. Idiom and phrasal verbs do not represent the analyzed data to draw the conclusion as there are only one idiom and four phrasal verbs out of twenty-six data (19%) found. 
 Keywords: source-oriented, target-oriented, machine translation, English-Indonesian, Instagram caption