Analyzing the Effects of Lexical Cognates on Translation Properties: A Multivariate Product and Process Based Approach

Abstract

AbstractThe translation of cognates has received renewed interest in Translation Process Research (Oster (2017) Empir Model Transl Interpret 7:23; Hansen-Schirra et al. (2017) Predicting cognate translation. In: Hansen-Schirra S, Czulo O, Hofmann S (eds) Empirical modelling of translation and interpreting. Language Science Press, Berlin, pp 3–22) but tends to be relatively time-consuming due to the manual identification of cognates and their translations. On the basis of work by Heilmann (Profiling effects of syntactic complexity in translation: a multi-method approach. PhD thesis, 2021) and the structure of the TPR-DB, we devised a relatively simple way to determine the cognate status of ST words and detect literal, cognate translations of cognates. We assess the “cognateness” of ST items on the basis of formal (dis)similarity with aligned ST and TT words. We use these measures to show how the cognate status of ST words and the literal cognate-to-cognate translation of cognates affect properties of the translation product and process. Using multivariate statistics, we are able to show that a ST token’s cognate status is a determining factor of translation ambiguity. We also find evidence for cognates affecting reading and typing behavior during translation. Additionally, we observe a moderation effect of translation experience on behavioral measures when cognates are translated literally. We interpret the results in support of the monitor model (Tirkkonen-Condit (2005) Meta 50(2):405–414) and propose a refinement of the operationalization of literality (cf. Schaeffer and Carl (2014a) Measuring the cognitive effort of literal translation processes. In: Proceedings of the EACL 2014 workshop on humans and computer–assisted translation, pp 29–37).