The paper addresses the use of corpora, both monolingual and parallel, in the
translation classroom. The languages used as examples are Italian and Slovene and the
specific feature dealt with is the translation of nominalization, which is both a
contrastive and translatological issue. After a short presentation of the theoretical
background and the reference corpus available for Slovene, as well as a parallel corpus of
Italian source texts and Slovene target texts, an example is shown of how certain issues
can be better resolved with the use of corpora instead of (or together with) the classical
bilingual dictionary. Monolingual corpora in the target language may show the
frequency and acceptability of a proposed translation equivalent, both as far as general
occurrence is concerned and regarding its distribution in different genres or text types.
Parallel corpora, on the other hand, provide evidence as to how problematic features
(even grammatical ones, as opposed to the frequently discussed lexical issues) are
translated in real translations, what options are available and which are more or less
acceptable within a given genre or text type.
