Corpora in translation: A Slovene perspective

Abstract

This paper reviews the use of corpora in translation practice and translator training, focusing on currently available monolingual and multilingual language resources for Slovene. The first part of the paper briefly outlines the state-of-the-art in corpus linguistics and then introduces publicly available corpora for Slovene, including general and special language corpora as well as several parallel corpora. The advantages and potential pitfalls of using the web as a corpus are also discussed. Part two presents some important considerations and guidelines for using corpora in both training situations and, more specifically, real-world translation projects. In many respects, corpora represent a richer source of information for a translator than dictionaries; on the other hand, a corpus user must know how to critically interpret the results obtained via a corpus query. Because corpora may not be readily available for many special domains and/or language pairs, procedures and tools for compiling one's own corpora are also described.