Translation Memory and Textuality: Some Implications

Much has been written about the ways in which technology has facilitated the rapid development of many aspects of the modern translation profession, including the re-use (‘leveraging’) of legacy data such as previous translations. This article adopts a different approach to the technology/translation relationship, posing a fundamental question, namely whether the use of computer-assisted translation tools—specifically, translation memory (TM)*—is changing the grammar of text. The question is approached through an analysis of TM data in a number of languages with a focus on cohesion between text segments, starting from the assumption that the greater the number of cohesive ties between segments, the lower the chances of re-use in new translations. It is concluded that it is problematic to isolate possible TM effects from expected translation shifts, either for system-related or genre-related reasons. Suggestions are made for the design of a larger study in order to resolve such difficulties