Translation-theory-driven prompts improve Spanish-Chinese journalistic translation quality

A study examines how prompt language and translation theory-driven prompt design influence the quality of Spanish-Chinese journalistic translations generated by GPT-5.2.

• The experiment tested 48 conditions using four prompt types, three prompt languages, and four El Pais editorials.
• Automated metrics (BLEU and BERTScore-F1) identified the baseline prompt as the best-performing condition.
• Human evaluation ranked the brief-oriented prompt highest with an MQM score of 8.66 compared to 7.84 for the baseline.
• Translation theory-driven prompts selectively reduced Awkward style errors, while Unidiomatic style errors persisted across conditions.
• Prompt language had a negligible impact on translation quality under both evaluation paradigms.

These results indicate that translation theory-driven prompts can yield measurable quality gains under expert evaluation of journalistic translations.