Ambiguïtés lexicales et structurales dans la traduction assistée par l’ordinateur par rapport à la traduction humaine

Abstract

This work analyses the problem of ambiguity in translation faced by computer assisted translation (CAT) tools. Our work was carried out through the examination of the production of a CAT tool in relation to human translation based on the Interpretative theory. We also analyzed a similar work to identify gaps to be addressed in ours. The point of departure in the study was that the production of the CAT tool is characterized by linguistic errors which impede the correct transfer of meaning. It was found out that these errors were as a result of inherent linguistic problems owing to lexical and structural ambiguity. Analysis of the production of a CAT tool confirmed that the computer is faced with problems of ambiguity in translation. The computer is unable to translate correctly lexical and structural ambiguity, because in such cases, a word, a phrase or sentence has more than one meaning. The analysis revealed that the inability of the computer to translate correctly lexical and structural ambiguity was due to its lack of contextual and real-world knowledge. The observation of the corpus enabled us to make a glaring finding: that the problems caused by lexical and structural ambiguity to the computer are solved by professional human translators; what is really ambiguous for the machine is not necessarily the case for humans. The study therefore highlighted its implications and suggested ways that could be helpful to the professional translators who use CAT tools in their work in order for the former to make efficient use of the latter. Keywords: Computer assisted translation; human translation; lexical ambiguity; structural ambiguity.