Translation of veterinary texts: Post-editing for different target audiences amid MT limitations

Abstract

This research fundamentally examines how machine-generated translations of veterinary texts might be post-edited. Given the specialized terminology and knowledge required in veterinary texts, their translations must be conducted by an expert translator with a background in translations of veterinary texts or medical science. Proficiency in translation technologies is also essential for an effective post-editing process. Predictably, the primary audience for veterinary texts consists mainly of professionals in veterinary medicine, including academics and veterinarians. Nevertheless, additional target groups, including an intermediate readership, i.e. veterinary students, pet owners or even farmers without veterinary expertise (laymen), still seek information for specific circumstances. In this study, a specialized text from the Merck Veterinary Manual or MCV (Abdul-Aziz et al., 2016) was randomly selected, and the raw translation output was generated using Google Translate. Analysis revealed that Google Translate's initial translations were done in a sophisticated language, primarily targeting professionals. Based on this finding, it can be said that despite utilizing neural MT systems, Google Translate may tend to overlook the potential variability of other target audiences or groups. Therefore, unlike a human translator, who likely possesses contextual knowledge in advance, Google Translate might fail to adapt translations for distinct audience groups. The fact that the raw outputs of the MCV from Google Translate adhere to a semantic or literary translation, primarily targeting professionals, suggests that light post-editing may suffice for this audience most of the time because they possess prior knowledge or experience in veterinary medicine. In contrast, non-experts may require full post-editing to thoroughly comprehend the text.