The aim of this paper is to design a model for literary text analysis which enables the identification and prediction of problems of translation. For this purposes two approaches are combined, viz. linguistics and literary studies. It is a descriptive and explanatory study which attempts to add accuracy and systematicity to a very complex field.