A Comparison of Translation Shifts in Jokowi’s Instagram and Twitter Accounts Using Machine Translation

Abstract

Translation shifts occur when content is translated from one language to another, due to either target language structure constraints (servitude) or translators' choices (options). With the growing use of machine translation services like Instagram Translate and Twitter Auto Translate on social media, understanding these shifts is important for effective cross-cultural communication. This study aims to compare translation shifts in the English translations of Jokowi's Instagram and Twitter accounts, focusing in the performance of these machine translation services. The research will analyze a sample of posts from Jokowi's Instagram and Twitter accounts, translated by Instagram Translate and Twitter Auto Translate. The study will use a combination of quantitative and qualitative analysis to identify different types of shifts, such as level shifts and category shifts, and to determine whether they result from servitude or optional choices made by the translation algorithms. Additionally, the research will assess the accuracy and idiomatic quality of the translations produced by these services, considering potential challenges related to context, idiomatic expressions, and cultural references. The findings will help improve understanding of machine translation performance in social media contexts and inform future improvements to translation algorithms. By examining translation shifts in Jokowi's Instagram and Twitter accounts, this study seeks to enhance communication and understanding on social media platforms across language barriers. The research also aims to provide valuable insights for users of machine translation services, particularly within the rapidly changing domain of social media platforms.