How relevant is the distinction between the original and translated texts in general corpora?

Abstract

General corpora do not always allow an average user to see whether the texts are written by native speakers or are translations, whereas it would be especially convenient to know whether the output of corpus searches reflects language use in originals, translations or a mixture of both. To show the relevance of distinction between the originals and translations, this work aims to investigate how the findings about a particular aspect of language may differ in original and translated Lithuanian. The sub-corpora of original and translated fiction have been retrieved from the Corpus of Contemporary Lithuanian Language and the Parallel Corpus. Corpus-based methods are used to analyse the general characteristics, the word frequency lists and pronoun usage in the two sets of data.