Abstract

The contribution presents a generative grammar for Greimas' metalanguage, designed to represent narrative competence, and investigates the possible automata to represent narrative performance, recognizing those narrative enunciates that belong to the grammar. In particular: (1) The grammar will generate both the narrative enunciates and their structural descriptions; (2) This way, the articulation of the immanent narrative structure will become clear; (3) It will become possible to study narrative features which are not manifested at the linear surface, such as recursivity. To test the model we will apply it to two different translations of the tale of the creation at the beginning of the Bible, thus demonstrating how, depending on the selected solution, the narrative structure changes. The model is a contribution to computational narratology, aiming to merge post-structuralist narratology with automata theory.