Dictionary, corpus and CAT tool use in legal and business translation: a comparative pilot study

This article aims at assessing the impact of the use of monolingual original corpora in the target language, hereafter MOC, containing texts produced by native speakers, on translation quality with respect to errors related to the relationship between source and target text (adequacy errors) and errors in the target text independent of the source text (acceptability errors) (cf. (Daems et al 2014:62). Past limited research has shown that MOC have a positive impact on, among other things, text subject understanding, term choice and idiomaticity (Bowker, 1998) and that MOC raise language awareness (Bowker, 1999). Two experiments were conducted among a total of 56 students who used various translation aids (bilingual dictionary, translation memory (TM), self-compiled MOC, online resources) for the translation of specialized (legal and business) text fragments. After determining the effect of the different translation aids on the number and the type of errors (established through error typologies) the positive effect of MOC on the error rate could be confirmed. TMs, whether or not combined with MOC, appear to be more oriented towards adequacy. TMs in combination with MOC lower the number of adequacy errors with regard to word sense disambiguation (hereafter WSD): TM use increases the number of context-correct translations compared to dictionary use. MOC could also have a positive effect on the number of acceptability errors (e.g. idiomaticity and grammar), as MOC use lowers the number of acceptability errors in our experiments. However, this positive effect may also have been induced by the combined use of MOC, dictionaries and online resources, as the students were allowed to use MOC next to other resources.

Jaar van publicatie:2022

Toegankelijkheid:Open