Intertextual Interpretative Semantics.

Anthropocentric computer assistance for text comprehension.

 

Abstract

Understanding natural language is usually considered as a data processing task. Such a conception rules out the interpretative competence required for text comprehension. Basing our approach on F.Rastier's theory of Interpretative Semantics ( IS), we propose a computer framework to assist text comprehension (called Intertextual Interpretative Semantics -- IIS) which differs from the data processing paradigm (e.g. Natural Language Processing) since it recognises the fundamental role of man in producing meaning. This is actually achieved by integrating man into the application which is consequently called anthropocentric. Furthermore, IIS extends IS theory by considering intertext as an object of study. A text may only be interpreted after its insertion in a more global textual entity, called anagnose. Intertext, by means of the concept of anagnose, becomes synonymous with the textual universe established by a reader for a given analysis of one or several texts. On the other hand, IIS integrates the notion of semantic sources (interpretants) in the definition of the semantic class. Consequently, intertextuality also becomes a relation connecting the class with its interpretants, and is thus located at the heart of the interpretation process. With the aid of a multi-user application framework, designed in UML and programmed in Java, the IIS postulates that the textualisation of context together with an appropriate organisation of this textual quasi-world enables the computer system to efficiently reuse the semantic material in order to propose interpretation assistance in a set of domains which demand a certain textual practice, like literary analysis and teaching, the elaboration of advertisements, translation, etc.