Aarhus University Seal

Panel 11. Translation revision

Revision of translations is an old activity but it has only very recently attracted the attention of any significant number of Translation Studies scholars, and courses devoted to revision within translation programs are still relatively rare. Yet developments in the world of translation would seem to call for more attention to revision.

As the use of Translation Memory and Machine Translation gradually increases, more translators are finding themselves checking machine outputs. And as volunteer translation expands, volunteer revision is also growing, as can be seen at Wikipedia’s English “proofreading” page, which in June 2015 listed some 900 volunteers for revision of translations of Wikipedia articles into English from 45 languages.

An interesting feature of revision is that there are very few machine aids specific to this aspect of translation work. No software can help the reviser detect unidiomatic word combinations, language that is too formal or technical for the intended readership, nonsense, deficiencies in inter-sentence connections, or most errors in transfer of meaning or in the focus structure of sentences. It seems that, for the time being, revision remains largely an activity of human minds unassisted by machines, while the drafting work of translators is on the contrary increasingly machine-assisted.

Revision of other people’s translations is also an interesting site of conflict between professional and business concerns: revision seeks to create adequate quality but it takes time, and therefore—unless the time for the drafting phase can be reduced—it increases costs. This situation gives rise to an ethical question: to what and to whom will the reviser be loyal?

Existing work by TS scholars has brought to light some interesting questions for research. First, this work has shown that, whether they are revising their own or someone else’s translation, different translators use different methods. Does the method have any effect on speed or on the quality of the output? The answer is still very unclear. Second, existing work has revealed problems: revisers failing to detect errors in translations, introducing errors, or wasting time on unnecessary changes. More generally, little is known about the usefulness of revision: how many problems in translations—especially serious problems—are being corrected (or not corrected!) per hour of revision effort in translation services. Studies of revision may at some point be able to help here if they shed light on the causes of these problems or lead to improved revision training.

Finally, because revision is mostly a reading rather than a writing process (its purpose is to spot problems in the draft translation), the study of revision has potential to focus on how translators read, whereas most studies in our field concern how they write.

The recent increased interest in revision within TS may have been in part triggered by the publication in 2006 of the European standard for translation services EN 15038, which requires translation providers certified under it to have every translation revised by a second translator (a requirement which has been carried forward to the new international standard ISO 17100 published in 2015). More research is needed on the revision policies of translation providers and how these policies are reflected in actual workplace practices.

Contributions are invited on the following (or other relevant) topics:

  • empirical studies of self-revision or other-revision
  • Machine Translation post-editing
  • impact of Translation Memory on translators’ revision work
  • content and pedagogical methods for revision courses and workshops
  • revision policies of translation services
  • workplace revision practices
  • application of the revision requirements of published translation standards
  • multilingual revision terminology
  • relationships among translation revision, monolingual editing and rewriting of specialised texts for lay readerships.
  • revision by volunteers
  • loyalties of revisers.


  • Name(s) of convener(s): Brian Mossop
  • Affiliation: York University, Toronto, Canada
  • Email address: brmossop@yorku.ca