A Grammatical Error Analysis Produced by Google Translate

Grammar has an important role to build a communication process, it is used to make a clear statement about the time of a verb occurs. This research is aimed to find out grammatical errors in terms of tense, active, and passive voice abstracts in translation of abtracts from Indonesian language to English produced by Google Translate. Descriptive qualitative method was used as the research method in this research. The population of this study was twenty abstracts that were taken from the University of Bengkulu’s Repository. The instruments that was used to collect the data were documentations and checklists. Keshavarzʼs (1999) theory was applied to analyze the data. From the data analysis it was found that tense was the most dominant error produced by Google Translate. Meanwhile, the active and passive voices error were rarely found. It can be inferred that as a tool, Google Translate cannot identify the tense of the sentences being translated. This is possible because tenses are closely related to the context of which the time a verb in a sentence happens, which can only be analyzed and understood by humans