Critical MT praxis: Machine translation and tourism slogans in Latin America

Abstract

Tourism slogans are not just advertising. They are cultural invitations, often the first words that international travelers encounter. Increasingly, that encounter takes place through free machine translation (MT) tools such as Google Translate, DeepL, and Microsoft Translator. This pilot study asks whether those tools preserve not only meaning but also the emotional tone that gives slogans persuasive force. To investigate, we introduce Critical MT Praxis, a framework that combines Baker's taxonomy of translation shifts with an affective semiotic lens. A set of forty official Latin American tourism slogans was analyzed across three machine translation engines and compared with human benchmark translations. Each output was assessed for fidelity, translation shifts, and persuasive effect using a five-point scale. The results revealed a sharper divide than anticipated. In just over half of the slogans (52.5%) maintained high fidelity, while the rest showed common shortcomings, including flattened tone (25%), rigid literalism, and culturally misleading cues. DeepL consistently produced the most natural and faithful renderings. Google and Microsoft split performance, often reducing tone or translating idioms too literally. Statistical patterns suggest that tonal drift and lexical inaccuracies most often erode persuasive impact. The findings underscore a key point. In tourism marketing, accuracy alone is insufficient, translations must also preserve rhetorical force and emotional resonance. For practitioners, testing slogans in MT engines, revising brand-critical terms, and auditing for accessibility are essential to preserve trust, identity, and competitiveness.