A la recherche d’un modèle explicatif du sens en traduction

The present article focuses on the cognitive mechanisms of deverbalisation and reverbalisation of sense in translation. Through an analysis of three professional translators’ process-oriented on-line data, made up from their think-aloud protocols (TAPs) and Translog files, I set out to show that the understanding and the transfer of sense are cognitive activities based on the translators’ ability to visualize, i.e. to construct mental images, before attempting to find the most appropriate linguistic means to represent these mental images in the target language. The units of analysis retained here are metonymic and metaphoric expressions. They have in common the fact that each is a conceptual entity which gives rise to another conceptual entity. However, as it is difficult to differentiate them, metonymy often shading over into metaphor, while metaphor often emerges from metonymy, it seems convenient to treat them here as interactive entities, or rather as a metaphtonymic unity, as suggested by Goosens (2003 : 431).