Harnessing the lawless: using comparable corpora to find translation equivalents

Bilingual dictionaries provide basic translation equivalents for a headword and typically limit the set of equivalents to words of the same part of speech as the headword. However, words taken in their contexts can be translated in many more ways. At the same time, equivalents listed in dictionaries are not adequate in many contexts, because of the contextual and collocational sensitivity of target language expressions. The problem is particularly acute for novice translators who lack the experience for finding contextually-appropriate translations. The paper proposes a methodology for finding translation equivalents in comparable corpora. This helps in training translation students to be aware of the translation potential of polysemous words from the general lexicon.