A brain-state-in-a-box network for narrative comprehension and recall

Understanding a discourse is considered to involve a series of specific processing phases which final result is a complete semantic, mental representation. This result is not only a representation of the text, but rather of what the text is about. Argument overlapping, connections between text constituents, and situational continuity, connections between discourse segments due to causality, spatiality and temporality, are the two main factors which potentially influence the extent to which readers form such a coherent mental model. This article explores the cognitive representation of narrative prose and the relations among the ideas in text. In our simulations, sentences in narratives are represented as nodes in a brain-state-in-a-box (BSB) network. Competing coalitions of the nodes drive the network into a stable equilibration. Quasi mental clusters (QMCs) are then extracted from narratives. Results shown attest the psychological validity of the model. The high correlation of the QMCs to psychological models may suggest the formation of mental models in text comprehension