A Descriptive-Comparative Analysis of the Impact of Machine Translation on the Rhetorical Features of AI-Generated Political Speeches

This paper examines the impact and implications of relying on machine translation (specifically, Google Translate) in translating culturally and ideologically charged political speeches, which are characterized by their rhetorical and functional formulation strategies. This paper utilizes three professionally generated Arabic and English political speeches, employing precise commands, as a case study for this research. The Hatem and Mason (1990 and 1997) model is used to analyze political speech and rhetorical elements before and after machine translation. The study concludes that machine translation is grammatically efficient and ineffective in preserving rhetorical, cultural, and ideological shifts. The data is analyzed speech by speech, each with its machine translation, according to the elements of political discourse as outlined in Hatem and Mason's model. Professional human translations are added to make the comparison and analysis more theoretically complete and convincing. Finally, this study recommends further research and collaboration on the potential of machine translation and professional human translation when handling sensitive data, such as political speeches.