Creativity in translation: machine translation as a constraint for literary texts

This article presents the results of a study involving the translation of ashort story by Kurt Vonnegut from English to Catalan and Dutch using threemodalities: machine-translation (MT), post-editing (PE) and translation withoutaid (HT). Our aim is to explore creativity, understood to involve novelty andacceptability, from a quantitative perspective. The results show that HT hasthe highest creativity score, followed by PE, and lastly, MT, and this isunanimous from all reviewers. A neural MT system trained on literary data doesnot currently have the necessary capabilities for a creative translation; itrenders literal solutions to translation problems. More importantly, using MTto post-edit raw output constrains the creativity of translators, resulting ina poorer translation often not fit for publication, according to experts