This work examines a case study that investigates (1) the achievability of extracting typological features from Polish texts, and (2) their contrastive power to discriminate between machine-translated texts from English. The findings indicate potential for a proposed method that deals with the explainable prediction of the source language of translated texts.