The paper applies RT to analyse an 18th century translation of a Latin text by the preeminent Romanian scholar Demetrius Cantemir. The translation diverges significantly from the original and was met with harsh criticism. Using the conceptual toolkit of RT, I argue that the differences between the original and its English translation were motivated by the translator’s desire to yield the same cognitive effect without putting the audience to unnecessary processing effort. Both effects and effort need to be evaluated by taking into account the respective cognitive environments of the source-text and the target-text audiences. The intertextual dimension of the text under scrutiny adds to the difficulty of communicating the same message in different languages and cultures.