Editorial intervention in translation: An experimental approach

Abstract Before publication, a translation may go through various intermediate stages and involve external agents other than the author and translator, such as editors or proof-readers. The role of different agents in translation has attracted increased attention in recent years, and studies have found that editorial intervention significantly affects translated language: Bisiada (2018b) finds extensive editorial changes to nominalizations in translation manuscripts while Kruger (2017) finds evidence that translation properties might result from editing rather than translating. In this study, we used German translated texts that were produced in two lab experiments, reported in Heilmann et al. (2021) and Heilmann et al. (2022), and asked professional German editors to edit them. The original English texts that the translations were based on included two different kinds of stimuli, namely non-agentive constructions and of -NPs, which, for the most part, were translated rather literally by the German translators even though previous corpus analyses had revealed a variety of non-literal translation strategies. Changes made to these stimuli by the editors were tracked, categorized, and statistically analyzed with the help of linear mixed models. The analyses show that editors are in fact more likely to change the complex stimuli that remained mostly unchanged in the German translations. While the kinds of changes made by the editors did not mirror the corpus results perfectly, the frequencies of complex stimuli after the translations had been edited were very close to frequencies reported in corpus-based studies like Freiwald (2016).