The talk deals with optimal reference translations (ORT). ORT are translations thought to be the best possible that can be achieved by a team of human translators. As the quality of standard reference translations is commonly questioned, ORT can be used in assessments of excellent machine translations in further research.
The lecture focuses on the creation process of ORT, translation quality of “standard” “professional” translations of newspaper articles (including source-language interference and translationese), and ORT evaluation. ORT were assessed in comparison with “standard” translations by translators, translation students, and non-translators. Based on the results of our evaluation, we can confirm a significant quality increase and we also document the relationship between evaluation and translation editing.