By using corpus of technical, journalistic and literary texts, originally written in English and translated into Portuguese, we may analyse the prevailing strategies employed by the translators when dealing with similarities and differences of source and target texts, languages and cultures. With this purpose in mind, we applied the descriptive-comparative model suggested by Aubert (1984, 1998). This model adapts the categories proposed by Vinay and Darbelnet ([1958, 1977]1995), in order to first identify then classify the procedures or modalities that characterise the translation of these text genres. According to data comparison, the most common modalities in the three corpora are found to be literal translation, transposition, and modulation. Because the high frequency of literal translation appears in technical texts and, contrary to what was expected, also occurs in journalistic texts with even higher percentage, we may infer a tendency towards automatism in the translation of these two text types. On the other hand, the greater need for modulations and transpositions with modulation may be said to be related to the literary text translators' more active participation in attempting to escape from literalism towards re-elaboration, in order to convey features commonly attached to the language of the novel.
Technical translation; Journalistic translation; Literary translation; Corpus-based translation research
