U

The unit of translation has been considered from a variety of perspectives. One early treatment comes from the comparative stylistics of Vinay and Darbelnet (1958/1995), who define the translation unit as ‘the smallest segment of the utterance whose signs are linked in such a way that they should not be translated individually’ (1995: 352). Vinay and Darbelnet's approach has been criticized for being overly prescriptive, too focused on the source language and based on idealized translations, factors which limit its ability to account for much real-world translation (see, for example, Ballard 1997). Their approach has been largely superseded by more recent, empirical research in translation, although it remains influential in translation pedagogy (e.g. Jones 1997) and the issues that arise from their discussion of translation units remain current. These include: whether or not such units are units of the source language/text, whether they are semantic or syntactic, at what linguistic rank they are realized, whether they have any cognitive basis, and whether they are conventionalized to any significant extent.