Tradução
INTRODUCTION Translation, as indeed any speech act, of any nature or description, is something which occurs between and among individuals and social groups. Translation is also something which takes place between different cultures, ideologies and world images. Furthermore, translation is something which goes on the whole time on the marketplace, involving, in economic terms, an added-value of several US$ billion a year. Translation is, evidently, something which is done to texts and discourses. And last but, probably, not least, translation is something which expresses itself in sentences, phrases, words. It is my purpose in this paper to provide empirical evidence which demonstrates that, despite the relevance and, indeed, the compelling urgency of adequate investigations into all textual and extra-textual matters related to language in general and to translation specifically, there is still sufficient scope for a closer look into the actual phrase and sub-phrase linguistic mechanisms that manifest themselves in each and every translational act. Indeed, it would be reasonable to expect that the macrostructures revealed on the planes of discourse, text grammar, pragmatics and cultural insertion of texts and their translations in one way or another would be mirrored by the microstructure of sentences, phrases and words. But, if de Saussure’s concept of the signifiant/signifié relationship being of an arbitrary nature holds good, the challenge remains to determine the manner and the extent of such mirroring. Beyond the theoretical interest of the approach suggested in the preceding, a number of praxiological circumstances seem to back up the relevance of a descriptive analysis of sub-phrase trends in translation. Thus, the advances in machine-assisted translation over these last 10 or 15 years, and which to a large extent derive