Dealing with Speakers’ Errors and Speakers’ Repairs in Simultaneous Interpretation

AbstractThe literature on the translation of impromptu speech tends to focus on the question of whether this text type is easier or more difficult to translate than prepared speech. Existing evidence, however, does not allow for decisive conclusions. In this article the authors attempt to identify the strategies adopted by interpreters when dealing with speakers' errors and speakers' repairs in impromptu speech. The material consists of three live Dutch speeches and their simultaneous interpretation into English by two professional interpreters. The material was analyzed with the help of the coding scheme presented in Levelt's research on selfmonitoring and self-repair in speech. The analysis reveals that in more than 4 out of 5 cases the interpreters corrected the speakers' unrepaired errors and translated the speaker's repairs without translating the original utterance. There was no significant influence of the type of repair (error repair vs. appropriateness repair), the moment of the repair, or the presence of an editing expression. The manner of restarting a repair, on the other hand, appeared to make a significant difference: afresh start gave rise to more translation problems. In analyzing the errors and repairs produced by the interpreters themselves, it became clear that interpreters' peiformances display stylistic differences. Additional informationNotes on contributorsFred van BesienFred van Besien, Kortrijksesteenweg 420, B-9000 Gent, Belgium. fvbesien@vlekho.wenk.beChris MeulemanChris Meuleman, Twee Dreven 26, B-9830 Sint-Martens-Latem, Belgium. cmbl@pandora.be