Anotace:
This article aims to test the so-called unique items hypothesis on Czech language data. The hypothesis formulated by Sonja Tirkkonen-Condit presumes the underrepresentation of unique items (target-language elements that have no direct counterparts in the source language) in translated texts compared to non-translations in the same language. A monolingual comparable sample corpus consisting of Czech translations and non-translations served as language data, both subcorpora containing approximately 17 million tokens. The tested Czech unique items were chosen from lexical units, word-formation phenomena, syntactic structures and language use phenomena. The frequency of these items was subjected to statistical testing with R software. The results reveal a general tendency of translated Czech language to contain less unique items. However, some of the individual items do not correspond to this tendency.