TEXTUAL GENERICITY AND AUTOMATIC TRANSLATION

Abstract

Texts for translation are nowadays more and more numerous, while people making translations for professional purposes or otherwise do not always have training in the field of translation and traductology. Therefore, it is frequently that they resort to automatic translation. This paper will approach a new issue, that has little been studied so far, namely finding out the relationship between genre and text type, and automatic translation, respectively. Within a context where the text has a permanent dynamic, where we deal, beside the << classical>> texts that can be localized in time and space from the production viewpoint, with closed texts in terms of form, even if they are semantically so very open, and up to hypertexts, in which the form participates in the construction of the meaning, the criteria of textuality established by text linguistics should be revisited. Thus, there are two types of questions which appear: firstly, what approach should we have to the act of translation, how is it reported to the text and what types of translation operations are mobilized; secondly, how do the genre and type of text to be translated influence the translation. Based on the automatic translation of a text corpus belonging to various genres, this paper will study the conditions of appropriacy of the act of translation carried out by means of software in function of the text genericity. Similarly, we will analyze the relation of the translator resorting to automatic translation as a tool of support with the text being translated. The investigation will comprise scientific texts, as well as literary ones and texts specific to web patterns of communication.