Who's afraid of false friends? Cognate ratios in translated and non-translated Dutch

This paper presents an empirical study on the proportion of cognate words (cognate ratios) in translated Dutch texts, compared to cognate words ratio in texts originally written in Dutch. To this end, we compiled a gold standard with manually verified cognate pairs for both studied language pairs, viz. English–Dutch and French–Dutch. In this study, we propose three hypotheses about how translators deal with cognates: (1) translators use the high degree of formal and semantic overlap between cognate translations to their advantage so as to produce the “easiest and fastest” translation (default translation hypothesis), (2) the higher the level of cognateness between a source and target language, the higher the cognate ratio in translated texts will be (cognate facilitation effect), (3) the higher the level of cognateness between the two languages, the more translators will be hesitant to use cognate translations (fear of false friends hypothesis). The results show a mixed picture: whereas not much evidence has been found for the first two hypotheses (depending on the respective language pair), the third hypothesis was confirmed. Further evidence, however, is needed from other language pairs, as cognate-receptiveness appears to be language-specific.